ADAPTIVE FUNCTIONS OF FLESHY ORNAMENTATION IN
WILD TURKEYS AND RELATED BIRDS
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
"That's the whole problem with science.
You've got a bunch of empiricists
trying to describe things of unimaginable wonder."
My advisor, Jane Brockmann, has gone above and beyond the call of duty in helping
me with this project I will always be grateful for her enthusiasm, advice, instruction,
guidance, constructive criticism and support. Little did I know that an academicumwelt is
learned as well as innate, just like everything else.
This study would not have been possible without the help of the rest of my
committee. Ellis Greiner kindly taught me parasitology and allowed me to use his lab. He
has also been very tolerant of the strange questions we ask over on this side of Archer Rd.
Rich Kiltie has thought carefully about the theoretical underpinnings and techniques I have
used to study fleshy structures. He directed me at several crucial turning points and made
me realize the consequences of my assumptions (although I suspect he does not find
turkeys very aesthetically pleasing). Doug Levey has always encouraged and helped me by
considering even my wackiest ideas with tact He is also my secret source for everything
"birdy." Brian McNab taught me about physiological ecology and generously allowed me
to use his lab. Dave McDonald has served as a "ghost" committee member and a friend
during my entire graduate career.
Literally dozens of people helped catch, handle and process wild turkeys. I am
grateful for their help. This study would not have been possible without it. In particular I
would like to thank those who helped me several times, including Laurie Eberhardt Ron
Clouse, Paula Cushing, and Monica Marquez. Farol Tomson and Frank Nordlie found a
place to keep the turkeys, the Department of Zoology paid the costs of maintaining them,
and Chris Wilcox and her crew helped care for the turkeys when I was out of town. The
Florida Museum of Natural History, University Athletic Association and Frank Maturo
generously allowed me to use their facilities to house the turkeys. My dad, Eduard
Buchholz helped improve the turkey cages. Margaret Byrne and Gerry Ryan helped build
My studies of free-living wild turkeys were done under permit from the Florida
Department of Natural Resources and Florida Game and Freshwater Fish Commission.
Permits were facilitated by Dan Pearson and David Cobb. I thank them and the staffs at
Paynes Prairie State Preserve and Camp Blanding Wildlife Management Area, particularly
Jim Weimer and Jack Gillen. The National Wild Turkey Federation provided wild turkey
transport boxes free of charge.
Several pilot studies did not pan out in the long run. Nevertheless I would like to
thank those who assisted me in these preliminary efforts. John Eisenberg kindly provided
permission to conduct preliminary research at the Katharine Ordway Preserve. Peter and
Denise Buchholz figured out how to catch jungle fowl on Kauai. Steve Nesbitt taught me
all about sandhill cranes, shared his data and let me tag along on crane capturing trips.
Marilyn Spaulding, Don Forrester and Sam Telford helped with my earliest attempts at
parasitology. I hope they will continue to encourage students as they did me.
At times graduate school is a very frustrating and unrewarding place. I must thank
the family and friends who have supported, laughed and commiserated with me over the
years and made it better. All have been instrumental in keeping me (somewhat) sane and
(usually) happy. Nothing would have been possible without the love and generosity of my
parents, Eduard and Elisabeth Buchholz. I hope someday they understand why I had to do
this. Gratitude goes to their son Peter, who reaffirmed our brotherhood despite having
chosen a different path. Thanks go to Marianne von Osten and Susan Edelbach, my "aunt"
and "sister," for always welcoming me home. Special thanks go to fellow cohort member,
peer advisor and buddy Laurie Eberhardt. I will miss her (boo!) and Peter. Carlita
Restrepo is a kindred spirit. I thank the expatriates, Tes Toop and John Donald, for being
funny, stabilizing and never dull. They tell me that in Australia Occam's razor is used to cut
Gordian knots. Monica Marquez made sure I finished my NSF grant application and has
always been an encouraging friend. Margaret Byrne and Gerry Ryan have been kind to me
in many ways. I value my friendship with each of them and hope they figure out what they
are looking for. Renee Calarco and Barbara Crute helped me get here, helped keep me here,
and were always there. Maggie and Ted would not be happy about being mentioned in the
same sentence, but I cherished them both. John, Tom, Bubba and Jane taught me alot about
turkeys; may they rest in peace.
Last, but not least, I thank the office staffs: Carol "Sweet Cheeks" Binello, Alice
"Glamour Girl" McClaughry, Janet "Green Thumb" Zeigler, Lynda Everitt, Lori "Where
Are You Going?" Clark, Ruth Ann Czerenda, Evelyn Rockwell, Kenetha "I've Got my Eye
on You" Johnson, Tangelyn "You're Crazy" Mitchell and, of course, Gracie "Babes" Kiltie.
Also thanks to Pete in the stockroom. They kept life in Bartram and Carr interesting and
smoothed over the red tape. Keep up the great job!
My research was funded by the Department of Zoology at the University of Florida,
a Frank M. Chapman Memorial Fund Grant in Ornithology from the American Museum of
Natural History, a grant-in-aid of research from Sigma Xi, the Animal Behavior Fund of the
University of Florida Foundation, the Cracid Breeding and Conservation Center, and a
Threadgill Dissertation Fellowship from the College of Liberal Arts and Sciences at the
University of Florida.
TABLE OF CONTENTS
1 GENERAL INTRODUCTION 1
2 A COMPARATIVE ANALYSIS OF THE ADAPTIVE
SIGNIFICANCE OF FLESHY STRUCTURES IN THE
Hypotheses and Predictions 7
Inter-individual Assessment 7
Immediate benefits 9
Good genes models 10
Fisher's runaway selection 13
Fleshy structures as heat sinks 14
Solar collector 14
Inter-individual Assessment 18
Immediate benefits 18
Good genes models 18
Fleshy structures as heat sinks 22
Solar collector 22
3 ADAPTIVE FEMALE CHOICE FOR MALE FLESHY
ORNAMENTATION IN WILD TURKEYS 25
Study Species 30
Mate Choice Experiments 31
Live male experiment 32
Male model experiment 37
Correlates of Ornamentation in Wild Males 39
Live Males Experiment 40
Male Models Experiment 42
Correlates of Ornamentation in Wild Males 43
4 MALE DOMINANCE AND VARIATION IN FLESHY HEAD
ORNAMENTATION IN WILD TURKEYS 51
Live Males 54
Male Model Trials 56
Live Male Trials 57
Male Model Trials 62
5 THE THERMOREGULATORY ROLE OF THE UNFEATHERED
HEAD AND NECK IN MALE WILD TURKEYS 66
Subjects and Apparatuses 67
Experimental Design 69
GENERAL DISCUSSION 82
LIST OF REFERENCES 85
BIOGRAPHICAL SKETCH 94
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
ADAPTIVE FUNCTIONS OF FLESHY ORNAMENTATION
IN WILD TURKEYS AND RELATED BIRDS
Chairman: H. Jane Brockmann
Major Department: Zoology
Wattles and other types of unfeathered fleshy ornamentation in birds are puzzling to
evolutionary biologists because these structures increase the bird's susceptibility to blood-
feeding insects and heat loss, without providing any obvious benefits. This study tests
sexual and nonsexual hypotheses for the evolution and maintenance of fleshy
ornamentation in galliform birds, particularly the wild turkey (Meleagris gallopavo) .
A comparative study of the elaboration of fleshy structures in the avian order
Galliformes identified the evolutionary pressures under which fleshy structures may have
evolved. After controlling for the effects of body size and shared phylogeny, the lack of
clear cut associations between fleshy ornamentation and other types of ornaments suggests
that sexual selection has not played an important role in the evolution of these structures.
Strong negative correlations with latitude, on the other hand, are supportive of a
thermoregulatory role for these characters.
Sexual selection for the maintenance of fleshy ornamentation was tested in mate
choice and dominance trials using wild turkeys raised in captivity. If the wild turkey's
fleshy structures are maintained by female choice, females are expected to mate non-
randomly with respect to these characters. Females chose males with greater snood lengths
and skullcap widths. Males with longer snoods were more likely to be dominant in dyadic
encounters. In complementary field studies in north central Florida, male snood length was
found to be a "good genes" indicator of coccidia infection and body condition. Thus the
turkey's snood appears to be maintained by female choice for males who signal their genetic
The extensive vascularization of fleshy ornamentation suggests that it may function
in thermoregulation. By insulating the unfeathered bare heads and necks of wild turkeys, I
tested whether the fleshy ornamentation of this species is important in heat dissipation.
Insulated turkeys increased their oxygen consumption and experienced higher body
temperatures relative to uninsulated controls. These results concur with those of the
comparative study and suggest avian fleshy structures function in heat dissipation as well as
sexual selection. Nonsexual hypotheses for the maintenance of ornamentation in animals
are rarely proposed, but are necessary to understand the evolution of these costly characters.
Brightly colored plumage, large branching antlers and delicately curved horns,
elaborate tails and courtship behaviors in animals seem inconsistent with our general notion
of the parsimony of nature. These characters are often called "extravangant" or
"ornamental" because they are particularly striking in appearance to human observers and
detailed studies of their function have not been conducted. They are ornamental in the sense
that they make the bearer physically more obvious without having any direct, functional
purpose, such as in catching food items or digging burrows. These extravagant characters
may increase the predation risks of animals that have them and likely have costs in terms of
the energy needed to grow and maintain them. Despite the potentially high costs of having
extravagant ornamentation, little is known of the benefits. However the dramatic physical
appearance of these structures, coupled with their development at reproductive maturity or
seasonally with breeding has led most authors to suggest that they have evolved via sexual
selection, that is that they have some function in increasing mating success. Thus the high
costs associated with ornamentation are assumed to be counterbalanced by increased
reproductive success. Assumptions about the adaptive benefit of "ornamental" structures
under sexual selection often make intuitive sense. Unfortunately additional hypotheses that
might explain how these traits evolved over time or how they are currently maintained are
rarely considered. The goal of my doctoral research is to propose and test both sexual and
nonsexual hypotheses for the evolution and maintenance of a subset of avian ornamentation
commonly called fleshy structures (FS).
Combs, ceres, wattles and similar fleshy structures are widespread in the class Aves,
but are particularly common in the order Galliformes (quails, pheasants, grouse, curassows,
guans, megapodes). Despite the frequency of these fleshy characters, there has never been a
thorough examination of their adaptive significance. First, I conduct a comparative analysis
of the morphological and ecological correlates of FS in the Galliformes to test four
alternative hypotheses for the evolution of these anomalous structures (Chapter 2). These
hypotheses fall into two functional groups that are not necessarily mutually exclusive: (1)
inter-individual assessment and (2) thermoregulation. After examining the interspecific
correlates of fleshy ornamentation in this avian order, I evaluate these hypotheses in a local
galliform, the wild turkey (Meleagris gallopavo) . in subsequent chapters.
Male wild turkeys are perhaps the most ornamented birds in North America. They
have iridescent feathers, a hair-like beard projecting from their chest and most notably an
unfeathered, bright red, white and blue head and neck. On the head and neck are three types
of extravagant fleshy projections. Perhaps the most remarkable is the distensible frontal
process or snood hanging over the bird's bill. At the base of the neck are large fleshy
protuberances called frontal caruncles and on the side of the neck are small polyp-like
fleshy structures called side caruncles. During display these fleshy structures change
markedly; the snood distends and the caruncles are flushed with blood and become bright
red. The association of this change with courtship suggests that these fleshy structures
affect mating success.
Sexual selection is divided into two subsets: female choice and male-male
competition for access to females. If female choice is an important selective factor
maintaining male fleshy ornamentation in wild turkeys, variability in male ornamentation
should affect male mating success (Chapter 3). Similarly if male-male competition is a
significant selective pressure, variability in fleshy ornamentation should affect the outcome
of male-male interactions (Chapter 4). Two general models of female choice for male
fleshy structures are tested: "good genes" and "arbitrary preferences." Both types predict
that females will exert directional selection for the more exaggerated form of male
ornamentation. Good genes models propose that the character in question is assessed by
females because it is an indicator of the heritable vigor of the bearer, but the arbitrary
preference model predicts no such association between ornament and male viability. To
investigate how wild turkey ornamentation may have evolved by female choice, the
assumptions and predictions of these models must be tested. First, do females mate
randomly with respect to male traits, and if not, which traits do they assess in prospective
mates? Second, are the characters assessed by females particularly good indicators of the
male's success in dealing with factors limiting the fitness of wild turkeys? One factor
limiting fitness of males is their ability to displace other males away from females during
the reproductive season.
Wild turkeys are a highly polygynous species in which males attempt to restrict the
access of other males to groups of females during the mating season. Thus agonistic
interactions and social dominance may be important determinants of reproductive success.
Females may prefer to mate with dominant males if components of dominance are heritable
and will be passed on to their offspring. Subordinate males should avoid dangerous battles
with dominant males if they have tittle chance of winning. Fleshy ornamentation may play a
role in mutual assessment before male battles, because these structures appear to be good
indicators of condition and testosterone levels in related galliform species. To see if this is
the case in wild turkeys, the value of fleshy ornamentation in indicating the outcome of
male-male combat is assessed. Then I ask whether males actually judge the ornamentation
of other males before interacting with them.
Female choice and male-male competition are not the only hypotheses that might
explain the maintenance of fleshy ornamentation in wild turkeys. Fleshy ornamentation is
highly vascularized and may be used in thermal balance, particularly at high temperatures.
The value of unfeathered ornamentation in dissipating heat at high temperatures and the cost
of an uninsulated head and neck at low temperatures is assessed by experimentally
reinsulating the heads and necks of wild turkeys as though they were feathered (Chapter 5).
Ornamental structures and elaborate courtship behavior patterns in animals have
long puzzled evolutionary biologists. Recent work on these characters, which are often
quite bizarre in appearance, suggests that these evolutionary puzzles are not always
attributable to sexual selection. Natural selection can mold the specific characteristics of
sexually selected characters to minimize their costs, and it can select for structures that share
characteristics with sexually selected traits, but that have very practical functions outside of
mate choice and male-male competition. The goal of this study is to understand the
evolution and maintenance of fleshy ornamentation in galliform birds.
A COMPARATIVE ANALYSIS OF THE ADAPTTVE SIGNIFICANCE OF "FLESHY"
STRUCTURES IN THE GALLIFORMES
Combs, ceres, wattles, knobs, snoods and other fleshy structures are a virtually
unstudied evolutionary puzzle in avian biology. These areas of bare skin, often called
"fleshy" structures, are present in some members of at least 30% of bird families. In some
small families, fleshy structures are found in every species (e.g., Casuariidae, Fregatidae,
Callaeidae); in other larger families, fleshy structures are nearly fixed (e.g., Gruidae and
Cracidae) or widespread (e.g., Phasianidae and Meliphagidae). Fleshy structures appear to
be completely absent in only 1 1 of the 27 avian orders (Table 2-1). Oddly, despite the
frequency of these fleshy characters and their importance to avian taxonomy, there has never
been a thorough examination of their adaptive significance.
FS are generally assumed to be sexual ornamentation (e.g., Welty 1975) and recent
empirical studies provide strong support for both intersexual and intrasexual explanations
for the maintenance of these structures (Brodsky 1988; Boyce 1990; Hillgarth 1990; Ligon
et al. 1990; Zuk et al. 1990a; 1990b; 1992; Holder & Montgomerie 1993). Unfortunately
none of these investigations has addressed nonsexual functions for these structures. In
addition no studies have addressed the costs inherent to having exposed skin. FS are a
chink in the protective armor (feathers and scales) covering a bird's body surface. Exposing
bare skin to a world of temperature extremes, aggressive competitors and disease-carrying
insects would seem to be maladaptive. Have birds evolved these seemingly costly characters
solely in the context of increased mating success?
Table 2-1. Fleshy structures are found in 47 families in 1 1 orders in the Class Aves.
• Falconifo nines
Classification follows Morony et al. (1975). Field guides were examined to identify taxa with FS.
All species were not examined, tbus additional taxa may possess FS. Famines without annotation
lack fleshy structures; • order contains at least one species with a fleshy structure; A no fleshy
structures present in representative species surveyed in each family of order; * family contains at
least one species that possesses a fleshy structure.
Although it is ordinarily not possible to know the selective pressures under which any
trait evolved in the past, comparative analyses of the environmental, morphological and
ecological correlates of traits across species provide clues to the selective factors associated
with their evolution (Clutton-Brock & Harvey 1984). I conducted such an analysis to test
four alternative hypotheses for the evolution of FS in the Galliformes (quails, pheasants,
grouse, curassows, guans, megapodes). Galliformes are well suited for such a study
because their behavior, ecology and distributions are fairly well documented. They exhibit a
diversity of FS that are familiar to anyone who has seen a domestic chicken or turkey.
FS are not fleshy, that is, they contain very little muscle tissue. As used in common
parlance, however, "fleshy" accurately describes the soft or pliable nature of FS. Fleshy
structures (FS) vary greatly among taxa in size, shape, color and location and in their use
during display (Figure 2-1). Depending on the species, intraspecific variation in the color,
shape or presence of FS can be associated with sexual maturity (Hollett et al. 1984),
seasonal changes (Stokkan 1979), age (Buchholz 1991) and changes in motivation
(Johnsgard 1983; Franklin & Menkhorst 1988; Reichholf 1988). Consistent with the
extensive variation in the external morphology of FS is a wide range of hypotheses in the
literature on the adaptive functions of exposed skin in birds. These hypotheses generally
fall into two major, functional groups that are not necessarily mutually exclusive: (1) inter-
individual assessment and (2) thermoregulation. In the next section, I describe these
hypotheses and develop predictions for testing and distinguishing among them.
Hypotheses and Predictions
The apparent association of FS with sexual maturity suggests that they are an important
form of sexual ornamentation. Several hypotheses suggest that sexual ornamentation
functions in inter-individual assessment (i.e. an individual evaluates the ornaments of
Figure 2- 1 . Diversity of fleshy structure size, shape and change during display in the
Galliformes. Unfeathered areas (FS) bordered by stippling, a) and b) bare orbit and eye
wattle of male Syrmaticus pheasant before and during display, respectively; c) bare orbit
and dewlap of Pipile guan; d) bare head, neck and collar wattle of brush turkey, Alectura : e)
bare orbit of male wood partridge, Rollulus : f) knob, cere and wattles of a male curassow,
Crax : g) eye combs of male black grouse, Tetrao , during display; h) and i) cere and cere,
horns and throat lappet in male Tragopan before and during display, respectively.
another individual and this information determines whether or how they interact).
Traditionally, sexual selection theory has been divided into two subdisciplines: female
choice and male-male competition. FS, then, may be used for assessment both by females
and by other males. Females may assess a male's parental ability, his ability to provide
access to resources, his genetically based viability, or other features (Johnson & Mazluff
1990). Similarly, males may use the same or different characters to assess other males
prior to aggressive interactions (Borgia 1979; McKinney et al. 1990), or rarely, females. A
true test of these hypotheses requires standardized studies of the mate choice and
competitive behavior of each species. Instead I have assumed that if FS are subject to
assessment, they will co-occur with other characters that have been shown to be subject to
assessment in other species. All inter-individual assessment hypotheses assume that
females prefer males with the most exaggerated FS and that males will not fight as readily
with such males.
A male's epigamic traits may reveal direct benefits to the female such as his ability to
provide parental care, to feed the female during courtship, or his dominance over good
foraging areas. Female choice on these characters is favored because she and her offspring
benefit directly, rather than receiving indirect benefits from increased survivorship under the
good genes models discussed below (Kirkpatrick 1987). Jf FS have evolved to signal a
male's ability to provide immediate benefits, then the following predictions can be made.
(i) Fleshy structures will be most common and largest in breeding systems where males
most commonly provide immediate benefits to the female and her young (i.e. in
(ii) Because males and females have similar variance in mating success in monogamous
systems (Payne 1984), making mutual assessment more likely, the sexes can be expected to
be more similar in FS size than in non-monogamous species.
Good genes models
If differences in general viability (e.g., health, dominance) are due to genotypic
differences, females may be able to choose these "good" genes for their offspring
(Andersson 1986; Heisler et al. 1987). Female choice for good genes may be especially
important in systems where the male provides no care or other resources for the female or
young. In this case traits that honestly advertise male condition will be chosen by females
(Kodric-Brown & Brown 1984). The following hypotheses address specific viability
correlates that females may use to assess males (and that other males may avoid in
Parasite resistance . Hamilton and Zuk (1982) proposed that FS in birds are utilized in
the assessment of blood parasite burden in prospective mates. Avian blood parasites may
reduce fecundity (Korpimaki et al. 1993) or survivorship (Atkinson & van Riper 1991;
Bennett et al. 1993). At some stage in their life cycle, hematozoa lyse the blood or muscle
cell they occupy and release merozoites (asexual sister cells that subsequendy infect other
blood cells or tissues) or gametocytes (which are picked up by insect vectors during blood
feeding). Red blood cell lysis or muscle impairment may result in anemia that is detectable
in the color or turgidity of highly vascularized FS.
The Hamilton-Zuk hypothesis can be expanded to include other parasites or parasitic
mechanisms that may affect the size or color of FS. For example, intestinal coccidia
infection is associated with reduced blood carotenoid levels (Ruff et al. 1974) and may
affect FS pigmentation. Ligon et al. (1990) and Zuk et al. (1990b) found that infection with
the intestinal nematode Ascaridia reduced comb size in male junglefowl and that females
preferred males with larger combs (above a threshold size), and big combed males were
more successful in male-male contests. Hillgarth (1990) found a negative correlation
between coccidia burden and mating success in ringneck pheasants and showed that males
that extended their face wattles fully and for long periods had higher mating success
(although she does not mention a direct correlation between wattle size and parasite burden).
Ectoparasites may also affect ornamentation and mating success. Spurrier (1991) applied
pigment marks resembling hematomas caused by lice on the cervical air sacs of Sage
Grouse and found that females preferred the unblemished males. Unfortunately the only
measures of parasite prevalence available for most galliform species is of the blood
parasites, therefore I restrict my predictions to this group.
(i) Species with FS should be more susceptible to parasitic infection than non-FS
species (Hamilton & Zuk 1982; Read 1987). The unidirectional nature of this prediction
may seem counter intuitive. However, since female choice for resistant males may result in
higher mean resistance in the next generation, the co-evolutionary nature of host-parasite
interactions insures that over long periods of evolutionary time, the host will continue to be
susceptible to the parasite. This is a result of adaptations by the parasite that counter the
host's new defenses.
(ii) Fleshy structures should be larger, more sexually dimorphic and show greater
change during display in species with higher prevalences of hematozoa.
Plumage brightness . Female preference for brighter or more boldly plumaged males is
well documented in the avian literature (Burley et al. 1982; Zuk et al. 1992; Johnson et al.
1993; Hill 1994). Brighter males may have access to more or better quality food, be less
susceptible to parasites or be more dominant Alternatively females may merely prefer more
brightly colored males with no naturally selected advantage, sensu Fisher (Fisher 1958). In
any case, if females are choosing males based on their ornamentation, an association
between plumage brightness and FS presence and size might be expected.
(i) Species with FS should have brighter plumage than those lacking FS. Also FS size
should be larger in brighter species.
(ii) Species with FS should be more likely to have other types of ornamentation
associated with reproductive displays, such as crests and elaborated tails.
(iii) If sexual dimorphism in FS size can be attributed to the same selective pressure as
sexual dichromatism, namely female choice, then these types of dimorphism should be
positively correlated with one another.
Dominance interactions . If social dominance provides access to resources, females that
choose dominant males may give their offspring a genetic advantage (as well as direct
benefits in systems with paternal care or territoriality). If male-male competition selects for
the advertisement of male quality via FS ornamentation (Borgia 1979), females merely need
to select the male with the best FS. How might FS signal male dominance? Fleshy
structure growth seems to be mediated by testosterone (references in Ligon et al. 1990).
Testosterone has been linked with aggression in many animals and is related to dominance
in some birds (Wingfield et al. 1987). Therefore dominant males with high testosterone
levels may have the biggest FS. For example, in ptarmigans (Lagopus spp.) male comb size
is correlated with dominance (Moss et al. 1979), testosterone injections increase comb size
(Stokkan 1979), and males with bigger combs have higher mating success (Brodsky 1988).
Similar correlative relationships between FS size, testosterone and dominance status are
found in red jungle fowl (Gallus gallus : references in Ligon et al. 1990). Presumably,
intrasexual selection is a precondition for female choice for male dominance. If FS have
evolved for intrasexual signalling the following predictions can be made.
(i) Across species FS should be associated with systems in which male-male
competition is most acute (i.e. in polygyny).
(ii) Fleshy structures should be positively associated with characteristics important to
male-male competition, such as sexual size dimorphism (Payne 1984) and the presence and
sexual dimorphism of weapons such as tarsal spurs (Ligon et al. 1990).
Other good genes models . The inherent costs of FS in terms of heat loss and insect
damage would suggest that these structures may serve as handicaps (Zahavi 1975; Grafen
1990a; 1990b). Handicapping traits make it more difficult for males to survive and by
doing so reveal to females the males heritable vigor. Although such traits reduce average
male survivorship, all the offspring of males who survive despite the handicap inherit their
fathers' good genes and the sons inherit their fathers' attractive ornament as well.
Unfortunately the handicap model makes no comparative predictions about FS.
Fisher's runaway selection
This process, called arbitrary by some authors (Kodric-Brown 1993), has no
comparative predictions (Heisler et al. 1987; Buchholz 1992) because the selective agent is
random female preference, unrelated to naturally selected fitness advantages and could occur
under many environmental conditions.
The high vascularity and exposed nature of FS suggests that they may have some
function in thermoregulation (Lucas & Stettenheim 1972; Crowe & Crowe 1979; Whittow
1986). Intuitively it seems that FS should increase both the evaporative and non-evaporative
(radiation, convection, conduction) components of heat loss. In fact, birds lose a large
amount of body heat from unfeathered or lightly feathered surfaces such as the legs
(Baudinette et al. 1976) and head (Hill et al. 1980). This heat loss may be detrimental to
individuals living in environments much colder than their body temperature (Whittow 1986;
Gray & Price 1988). In species living in hot environments, however, heat loss from the
head may be essential to general homeostasis. Specific vascular adaptations may facilitate
heat loss to protect heat sensitive tissues. For example, a counter current heat exchange
system around the eye, the rete mirable opthalmica , is thought to keep the brain cooler
than the deep body temperature in some birds (Whittow 1986). Alternatively,
vascularization near the exterior of the animal may aid in the collection of solar heat when
the surrounding air temperature is low (Henneman 1988). Thus FS could function in
thermoregulation in two ways: (a) heat dissipation and (b) heat absorbtion.
Fleshy structures as heat sinks
Crowe's (Crowe 1979; Crowe & Withers 1979b; Withers et al. 1980) studies of
behavior and FS variation in the helmeted guineafowl (Numida meleagris) show that wattle
size in this species is positively correlated with maximum daily temperature across its broad
geographic range. Biotelemetric measurements of the brain temperature of captive subjects
show that these birds maximally exposed FS in warm ambient temperatures and reduced
total thermal conductance by retracting them in cold temperatures. If FS have evolved for
this purpose, these predictions follow.
(i) Fleshy structures should be more likely to occur and FS size should be larger in
species living at lower altitudes and latitudes. As species ranges get further from the
equator they have less need for dissipating heat
(ii) Similarly sexual dimorphism in FS size should be reduced in species living at lower
altitudes and latitudes. As species ranges approach the equator, it becomes equally
important for both sexes to lose excess body heat.
(iii) Change in FS size during display should be more pronounced nearer the equator if
these structures are used to dissipate the excess body heat generated by energetically active
To maintain body temperature during cool conditions, FS may be used to absorb heat
from the environment instead of increasing metabolic activity (Burtt 1986). The predictions
of this hypothesis are the reverse of those for the Heat Sink hypothesis above.
(i) Fleshy structure occurrence and size should be positively correlated with upper
latitudinal and altitudinal limits of species.
(ii) Because it should be equally important for both sexes to collect radiation in colder
areas, sexual dimorphism in FS should be negatively correlated with altitude and latitude
(iii) Because intraspecific studies show that FS size changes only during sexual
displays and not outside of this context, it is unlikely that size change is important to
collecting solar radiation. Therefore there should be no relationship between FS size
change and latitude or altitude.
Despite the extreme variation in shape and usage of FS, they share certain attributes and
can be divided into three broad categories: (1) sparsely feathered thin skin without
distinctive coloring (e.g., the lores of hawks); (2) thickened bare skin with or without
distinctive coloring (e.g., the orbital ring in some parrots); and (3) projections of thickened,
brightly colored skin (e.g., a chicken's comb). Lumping these three types of structures
under a single heading may be somewhat unnatural, since bird species may have very
different reasons for exposing their skin. However, so little is known about the
morphology, associations and functions of these variable structures that I think this
phenomenon warrants an initial investigation at a broader level. Therefore, to identify
patterns in the function of exposed skin, I initially lump these structures.
Using literature accounts I collected information on the presence, size, sexual
dimorphism and size change during display of FS, and on the upper, lower, and mid
altitudinal and latitudinal limits, body size and sexual dimorphism, mating system, presence
of spurs, blood parasite prevalence, and plumage showiness (or "brightness," sensu
Hamilton & Zuk 1982) and sexual dichromatism of 279 species of Galliformes in three
families (Cracidae, Megapodiidae, Phasianidae). Fleshy structures and plumage
characteristics were determined from plates and photos in monographs (Delacour &
Amadon 1973; Johnsgard 1983; 1986; 1988), a variety of field guides, individual accounts
and personal observations. The FS size of resting males was ranked on a scale of 0-5
relative to head size. The ranking of FS size relative to head size in all species prevents any
bias in the results where the FS in question may have been used to establish the underlying
taxonomy. Sexual dimorphism in FS size was ranked on a scale of 0-2. Fleshy structure
size change during display was ranked on a scale of 0-5. Plumage brightness was ranked
on a scale of 0-5. Sexual size dimorphism was calculated as the average male mass divided
by the average female mass. Mating system was ranked on a scale of 1-3 based on the
degree of polygyny thought to commonly occur in each species. The average number of
spurs on males of each species (Davison 1985) was included in the analysis as a measure of
male-male competition. Sexual dimorphism in spur number was measured as the female
value subtracted from the male value. Maximum and minimum altitudinal limits were
collected from species accounts and latitudinal limits were estimated from range maps.
Blood parasite prevalences (proportion of sampled individuals found to be infected) were
collected from regional surveys or reviews (Oosthuizen & Markus 1967; 1969; Greiner et
al. 1975, Ashford et al. 1976; Bennett & Herman 1976; Wink & Bennett 1976; McClure et
al. 1978; White et al. 1978; Peirce 1981).
Recently a number of new cladistic methods that control for the nonindependence of
data points due to shared phylogeny have been developed for comparative studies (Harvey
& Pagel 1991). Many of these assume a "true" phylogeny of the organisms under study.
Unfortunately the phylogeny of the Galliformes does not meet this condition because it is
in constant flux (Randi et al. 1991) and has had very few molecular studies at lower branch
points. The comparative technique I used controls for the effects of shared phylogeny by
averaging measurements within nested taxonomic subsets (i.e. averaging the species within
each genus, then genera within tribes) and examining the correlations of these averages at
each of these taxonomic levels. Pagel and Harvey (1988) have suggested conducting
statistical analyses of these data at the single taxonomic level that subsumes the most
variance. In the gaUiform data set analyzed here, most of the variance in FS characteristics
is found at the tribe or family level. The decision rule proposed by Pagel and Harvey has
drawbacks when comparative associations are studied within one avian order because
sample sizes are considerably reduced at higher taxonomic levels. Statistics cannot be
employed with only three values at the family level and low sample sizes at the tribe level
carry with them a high risk of type II error; the inability to detect significant relationships
between variables when they exist. Therefore, I adopt a more exploratory approach and
present the associations found between variables at all three taxonomic levels considered:
species, genus and tribe. Correlational relationships found only at all lower taxonomic
levels (e.g., across species) may be artifacts of shared phylogeny rather than the products of
unique evolutionary events. Correlations at higher levels may not be significant because of
small sample size. However those correlations that retain statistical significance across all
taxonomic levels probably represent strong adaptive relationships.
The hypotheses presented for the evolution of FS were tested statistically in three ways.
First I compared taxa with FS to those lacking these structures. The occurrence of FS was
tested in a contingency table constructed from the number of taxa above and below the
median of the pertinent independent variable and the presence or absence of FS. Second the
factors associated with the elaboration of FS in species that had them were also tested in a
contingency table. Fleshy structure size and the independent variables were divided into two
groups at the median. Third the strength of association between three FS characteristics
(FS size, dimorphism and display change) and the independent variables were tested with
non-parametric correlation analysis (Spearman rank correlation). Sample sizes vary with
the availability of data for each variable. The data set lacked complete information for most
of the megapodes, and therefore these species are poorly represented by the results.
At the species level, FS size was strongly correlated with body mass. To remove
correlations between variables caused by their shared correlation with mass, residual values
generated by regressing each variable with body mass were used in all the analyses. This
approach assumes that the effect of natural or sexual selection on FS characteristics is direct
and not mediated by body size. It was not possible to test predictions relating to sexual size
dimorphism because this variable is calculated from a ratio with male mass as the
denominator. In the contingeny analyses of the occurrence of FS this resulted in about half
of the FS with a rank of one being included in the "FS absent" row. Finally, latitude and
altitude were positively correlated with one another and showed similar associations in most
tests, so for the sake of brevity results given only for latitude. Except where noted latitude
refers to the mid-point of the species' range.
FS were more common and more elaborate in non-monogamous mating systems.
However after the effects of body size were removed, there was no difference in the
occurrence or size of FS between monogamous and non-monogamous taxa (Table 2-2 and
2-3). Sexual dimorphism in FS size was significandy reduced in monogamous species.
Good genes models
Fleshy structures were not more common or larger in more showy species. However,
sexual dimorphism in FS size was positively correlated with sexual dichromatism (Table 2-
3). Species with elaborate tails were more likely to have FS at the tribe level. Fleshy
structures were not significandy more common in polygynously mating species, nor did FS
size increase with number of spurs or the degree of spur dimorphism. Although sexual
dimorphism in FS size at the species level was significandy greater in species with above
median spur dimorphism, this relationship did not persist at higher taxonomic levels. The
occurrence and elaboration of FS was positively correlated with sexual size dimorphism, but
this relationship is probably due the fact that more dimorphic species are also larger species.
Table 2-2. Occurrence of FS in galliforms relative to predictions of hypotheses for
the evolution of these structures.
Hypotheses and Species Level Genus Level Tribe Level
Predictions X 2 (N) . X 2 (N) U (N)
Tmm^iate Benefits 33+ (204) 0.0 _ (43) -2.3* (8)
FS more common in
Pa rasite Resistance q.3 (52) 3.1+ (20) "°- 9 ( 8 )
FS more common in
taxa with high
Plvftnaffft Brightness 2.4 (204) 3.2+ (43) °'° ^
FS more common in
brightly colored taxa
FS more common in 2.O (204) 0.0 (43) ~ 2 ' 3 * ^
taxa with elaborate
FS more common in q.O (202) 2.3 (42) "°* 3 ^
taxa with elaborate
FS more common in 3 3& + (204) 0.0 (43) ~ 2 -3 a ^
FS more common in 2 8 a+ (204) 0- 1 ( 48 ) ~ 1,2 ®
FS more common in 3 ( 43) -0.3 (8)
taxa sexually D -° ^ VH)
dimorphic for spurs
HsaLSink . 6.4* (48) -2-3* (8)
FS more common rn **./ \iyv>
taxa with lower
-7 1 nQtt 6.8** (48) -1.7 + (9)
Snlar Collector 2A ( 1% )
FS more common in
taxa with higher
+ 09>P>0.05; * 0.05>P>0.01; ** 0.01>P>0.001; a significant, but in opposite
direction; nearly significant, but in opposite direction. Probabilities are for two-tailed tests.
i — i
1 — 1
I — I
Fleshy structures as heat sinks.
Fleshy structure size was negatively correlated with latitude, and sexual dimorphism in
FS size was positively correlated with latitude; that is, FS got smaller and more sexually
dimorphic at higher latitudes (Tables 2-3). The degree of change in FS size during display
was predicted to be greatest nearest the equator; however, the opposite relationship was
obtained. Display change in FS size increased significantly with latitude.
Species with higher latitudinal limits were not more likely to have FS, nor did they have
smaller and more sexually dimorphic FS (Tables 2-2 and 2-3). In addition there was a
positive correlation between maximum latitude and FS size change during display at the
genus level. As the minimum latitudinal limits of a species increases, females become less
likely to have FS and males become more likely to reveal the full extent of their FS size
only during display.
Despite the many intraspecific studies demonstrating the importance of FS to inter-
individual assessment, a comparative analysis did not find much support for this hypothesis
as an explanation for the evolution of FS across galliform species. This result must be
interpreted with caution because the expected correlates of inter-individual assessment
processes (e.g., plumage brightness or spur number) are themselves only hypothesized
correlates of female choice and male-male competition based on intraspecific studies. A
true test of this hypothesis would necessitate standardized studies of the mate choice and
competitive behavior of each species. Instead I have assumed that if FS are subject to
assessment, they will co-occur with other characters that have been shown to be subject to
assessment in other species. The veracity of this assumption may depend on what is being
assessed (Sullivan 1994). Additionally it was not possible to include other characters that
may be correlated with female choice, such as intestinal parasite prevalences. Even though
FS size was not correlated with the other measures of ornamentation, sexual dimorphism in
FS size was positively correlated with plumage dimorphism. This finding suggests that the
factors controlling these variables are the same. Because plumage dimorphism is generally
attributed to female choice for male brightness, sexual dimorphism in FS size may also be
explained by this selective pressure.
FS characteristics were strongly associated with the latitudinal distributions of species,
even after controlling for body mass. Fleshy structures were more common, larger and
more similar between the sexes in species nearer the equator. These results are consistent
with a heat dissipation function for FS, in congruence with intraspecific studies (Crowe
1979; Crowe & Withers 1979; Withers & Crowe 1980). Alternatively the inverse
correlation between size and latitude might be the result of the increased thermoregulatory
costs of having bare skin in colder regions. The increase in display change at higher
latitudes is also consistent with a cold-limiting rather than a heat-dissipation explanation.
However if cold temperatures limit FS size across species, one might expect that FS
occurrence and elaboration would be most strongly correlated with the maximum latitudinal
limit rather than the minimum limit This is not the case, suggesting that FS actively
function in dissipating metabolic heat in species living in warm climates.
Studies of sexual selection often neglect to propose and test nonsexual hypotheses for
the maintenance of ornamentation in birds and other taxa. Balmford et al. (1993) have
suggested that an ignorance of aerodynamics has led evolutionary biologists to attribute
incorrectly all long bird tails to evolution by sexual selection. They argue that some types
of long tails are not burdensome nor aerodynamically costly and thus would not make good
indicators of male condition. Similarly characters that were assumed to be maintained
solely for social signalling, such as the bushy tails of ground squirrels (Bennett et al. 1984)
or the ultrasound emissions of some rodents (Thiessen 1979), or for quite mundane
purposes like feeding, such as the flattened bill of most ducks (Hagan & Heath 1980), may
also be maintained as a result of their thermoregulatory functions. Considering alternative
or congruent hypotheses for the maintenance of ornamental characters provides a better
understanding of how seemingly extravagant characters can be quite functional in a practical
sense. Unfortunately in the absence of an accurate, detailed phylogeny it is impossible to
determine if FS first evolved for heat dissipation or in sexual selection. Comparative studies
of the probable ancestral states of ornamentation and behavior have provided unique insight
into how sexually selected characters first come under assessment by females (Basolo
1990; Hill 1994). Similar studies of ornamental characters with both sexually selected and
naturally selected hypotheses in mind may provide a better understanding of how
extravagant characters can evolve despite their apparent costs.
Fleshy structures differ from other forms of ornamentation in many ways. They are
living integumentary outgrowths and thus have inherent costs that other nonliving forms of
ornamentation, such as big tails or long spurs, do not possess. Similarly the highly
vascularized nature of FS makes them more likely to be functional outside of the realm of
reproduction. Therefore, they are uniquely suited for identifying the conflicting forces of
sexual and natural selection that produce different suites of elaboration in separate
populations of the same or closely related species (Endler 1983; Kodric-Brown 1990). A
recognition among empiricists and theoreticians that ornamental structures may have
multiple beneficial functions, such as the apparent heat dissipation value of FS, the
thermoregulatory values of squirrel tails, ultrasonic emissions and duck bills, may advance
our understanding of the evolution of extravagant characters.
ADAPTIVE FEMALE CHOICE FOR
MALE FLESHY STRUCTURES IN WILD TURKEYS
The wild turkey (Meleagris gallopavo) is a familiar example of a sexually dimorphic
species exhibiting an array of extravagant behavioral and morphological characters that
serve no obvious useful function other than to attract mates. Males weigh more than twice
as much as females and have iridescent feathers, a large fan-like tail, a long hair-like beard
projecting from their chest, tarsal spurs and an unfeathered head and neck. The unfeathered
neck is covered with pebble-like red bumps called side caruncles, that may fuse to form
horizontal bands of bumpy skin (Figure 3-1). At the base of the front of the neck are three
or four roughly elliptical pillow-like outgrowths called frontal caruncles. The skullcap on
the crown of the head has thickened skin that ranges in color from red to white and light
blue and appears to change depending on the motivational state of the animal. Perhaps most
striking is a distensible process at the base of the upper bill called the snood. Finally a thin
dewlap stretches from beneath the lower mandible down to the frontal caruncles. This
extravagantly ornamented bird is uniquely suited to testing models of female choice because
it is a highly polygynous species and males associate with females only for mating and
never provide care for their offspring.
Although we cannot be sure by which selective process any one of the turkey's
many ornaments has evolved during the history of this species, the evolutionary history of a
character can be inferred by examining the selective pressures maintaining such characters
in the present By testing the mating preferences of captive female wild turkeys and
Figure 3-1. Unfeathered head ornamentation of mature male wild turkey: a) skullcap; b)
relaxed snood; c) dewlap; d) frontal caruncles; e) side caruncles.
examining the correlates of the preferred types of ornamentation in wild males, I investigate
how female choice may be maintaining the extravagant ornamentation of male wild turkeys.
Hypotheses of indirect benefits of female choice fall into two types: good genes and
arbitrary preferences. Both types predict that females will exert directional selection for the
more exaggerated form of male ornamentation (i.e. they prefer bigger, brighter, louder, etc.
forms of the trait). Good genes models propose that the character in question is assessed
by females because it is an indicator of the heritable vigor of the bearer. Male vigor or
viability is a measure of a male's ability to survive acute or chronic factors that limit fitness.
There are two principal types of good genes hypotheses. The handicap model (Grafen
1990a; Zahavi 1975) proposes that ornamentation reduces average male survival enabling
females to assess any individual male's ability to survive despite his burdensome ornament.
In this case females that choose to mate with more highly ornamented males have higher
fitness because of the increased survivorship of their female offspring, who carry the male's
superior survival genes but do not express the sex-limited genes for handicapping
ornamentation. Male offspring who have genes for the attractive handicap experience
increased mating success in addition to increased survivorship. In the other type of good
genes models, females are assessing male ornamentation because it specifically indicates a
second, less visible trait determining male fitness. Hamilton and Zuk (1982) propose that
females assess the quality of male ornamentation because these characters are particularly
good indicators of the parasite burden of the bearer. In the parasite-assessment model,
females benefit from choosing the more ornamented males because their offspring will
inherit the male's ability to avoid deleterious infection. In similar models, the male's
foraging success (Kodric-Brown 1989) or age (Manning 1985; 1989) is honestly indicated
by his ornamentation.
Arbitrary preference models, on the other hand, hypothesize that the character
reveals nothing of the abilities or heritable vigor of the male but is nonetheless preferred by
females. Fisher's (1958) theory of runaway selection proposed that an initial female
preference for a male character that came about by natural selection or genetic drift could
become genetically linked to the male trait, resulting in elaboration of the trait beyond its
original utility. Similarly other authors (Kirkpatrick 1987; Enquist & Arak 1993) suggest
that specific attributes of male characters, such as color, shape or frequency, elaborate over
evolutionary time because they more effectively stimulate existing biases in the sensory
system of the female. In these models male offspring of females exercising an arbitrary
preference have greater reproductive success due to the character they inherit Because the
arbitrary preference model is empirically distinguishable only by the absence of a
correlation between the male character and measures of male viability, it proves difficult to
test The arbitrary preference model remains a default or null hypothesis in studies where a
female preference is detected but the adaptive reason for the preference is not explained by
the good genes model (Buchholz 1991). The only way to distinguish between the good
genes and arbitrary models of female choice, when it is not possible to monitor the next
generation, is by examining the relationship between variation in ornamentation and
variation in male viability.
Extensive research on the factors limiting population size in wild turkeys suggests
three major selective pressures: diseases and parasites, predation, and weather (Dickson
1992). Population surveys have detected over 100 species of viral, bacterial, fungal,
protozoal, metazoal and arthropod parasites that may limit the fitness of turkeys (Davidson
& Wentworth 1992). Although the impact of most of these parasites on their hosts is
unknown, experimental studies using domestic or captive wild turkeys have suggested that a
poxvirus (Forrester 1991), a proventricular nematode (Hon et al. 1975; 1978), and blood
parasites (Forrester et al. 1974; 1980; Atkinson et al. 1988) have the most severe impact on
host survival, particularly of juveniles. High levels of predation on nests and on nesting or
brooding females is well documented (Miller & Leopold 1992). Finally weather conditions
in the northern areas of the wild turkey's range affect survivorship and fecundity.
Populations subject to deep snow can suffer widespread starvation and mortality
approaching 60% (Healy 1992b). Food limitation during winter is probably not limited to
the northern areas of the wild turkey's distribution. A twenty-fold increase in visits to
artificial feeders during Jan-March suggests that winter food may be limited in Florida as
well (Powell 1965). How might females detect males that are superior in their abilities to
cope with these factors?
The handicap model predicts that females will prefer ornaments that reduce male
survival. These might be characters that make males more subject to predation, invite
challenge from competitors or increase the risk of starvation. To test the handicap principle
of ornamentation, experimental manipulation of the quality of male ornamentation of wild
individuals would be necessary. Because a test of the handicap model is beyond the scope
of this study, I do not discuss the predictions of this hypothesis further. However anecdotal
evidence addressing this hypothesis is presented in the discussion.
Since parasites are a ubiquitous problem for wild turkeys, the Hamilton and Zuk
(1982) theory of mate choice as an adaptation against parasitism seems a plausible
explanation for the maintenance of FS in this species. If turkey hens are using male
ornamentation to assess parasite burden, the intensity of infection by the most deleterious
parasites should be inversely correlated with the quality of male ornamentation that females
are using to choose a mate. To be favored by females, these ornaments should be better
indicators of parasite burden than other anatomical structures, such as bill or tarsus length.
Adult male wild turkeys begin their energetic courtship displays in early spring,
often while snow remains on the ground, relying on their fat reserves for sustenance
(Pelham & Dickson 1992). If the ability to access food and store energy for surviving
winter snow storms is an important determinant of survival as previous research suggests,
females may use the quality of a male's display as an indication of stored energy reserves.
If this is the case, female choice should rely heavily on the most energetic parts of the male
display and on any other characters that might be indicative of fat reserves, such as body
Age is perhaps the most honest indicator of good genes. Males that survive despite
parasites, predators, and food shortages have demonstrated their superior ability merely by
remaining alive. Thus females may choose mates based on ornamentation that reliably
reveals male age. Although beard length and body weight are positively correlated with age,
spur length is thought to be the best indicator of yearly age classes in eastern wild turkeys
(M g. silvestris : Kelly 1975; Steffen et al. 1990). This suggests that females should choose
to mate with males with longer spurs.
The arbitrary preference models for the evolution of male ornamentation predict no
relationship between ornamentation and male survival or condition. If the characters chosen
by females are not related to any of the factors that limit turkey survival or fecundity, it
seems reasonable to conclude that they are maintained purely by the increased mating
success of the male without any direct increase in the reproductive success of the female. I
test these hypotheses for the maintenance of ornamentation in wild turkeys with
experimental data collected from captive mate-choice trials and correlational data collected
from wild caught males.
Day-old wild turkey poults were purchased in May 1991 from a gamefarm (L&L
Pheasantry, Hegins, PA, USA) whose large breeder flock continued to experience gene flow
from wild males (M g- silvestris ) until a decade ago. Poults were raised indoors under
heat lamps with gamebird starter feed (Purina Startena, 30% protein) and water provided M
libitum . After eight weeks the birds were transferred to a large (5.3m x 5.3m x 2.6m)
outdoor pen at the Florida Museum of Natural History's Special Projects Laboratory.
After 14 weeks of age, males were maintained separately from the females in two
visually isolated groups of 16 in sand-floored aviaries (5.3m x 5.3m x 2.6m). Females
were moved to a cement-floored pen (5m x 12m x 4m) that was visually isolated from the
males. To minimize wastage only as much feed (Purina Grower or Maintenance, 19 or 12%
protein) as could be consumed in 3-4 hours was provided daily, varying seasonally.
Additionally a variety of green forage (e.g., cut grass, bamboo leaves, etc.) was provided
twice a week. Water was available ad lib.
Very limited displaying observed during pilot studies in early 1992 was attributed to
suppression by dominant males in the relatively crowded, group housing. To alleviate this
problem all males were housed individually in cylindrical wire cages (1.3 m in diameter and
lm high) at 15 months of age. Black plastic dividers prevented males from physically
interacting with males in neighboring pens, although they could see and hear other males in
pens 4 m away. Each male was fed approximately 0.5 1 of breeder rations (Purina Layena,
20% protein) daily, along with occasional peanuts, wild bird seed, or carrot or apple slices.
Water was provided in 0.5 1 containers and was replaced daily.
Mate Choice Experiments
To investigate how wild turkey ornamentation may have evolved by female choice it
is necessary to test the assumptions and predictions of the good genes and arbitrary
preference models. First do females mate randomly with respect to male traits, and if not,
which traits do they assess in prospective mates? Second are the characters assessed by
females particularly good indicators of the male's success in dealing with factors limiting the
fitness of wild turkeys?
The mating preferences of female wild turkeys were tested in two experiments. In
the Live Male experiment, females were given a choice between two naturally displaying
males. The characteristics of males chosen by females were compared to the characteristics
of males not chosen by females to identify the characters correlated with female choice. The
Male Model experiment gave females a choice between two artificial males that differed
only in the amount of head ornamentation. The model experiment ensures that the character
thought to be assessed by females in the live male experiment is not merely the correlate of
another unmeasured character actually under assessment by females.
Live male experiment
Live males . Thirty-three, 20-month old males were exposed to an artificial
photoperiod of 14 hours for four weeks prior to and during the mate choice tests. Prior to
the mate choice trials, males were classified as "displayers" or "nondisplayers" based on
whether they strutted consistently during a half hour observation period on each day of the
week before the first trials. During the trials no "nondisplayers" became "displayers," but
the reverse was true. The 1 1 males classified as consistent displayers were used in the mate
choice trials. Seven ornamental characters were measured prior to the first trial. Relaxed
snood length was measured from the point of attachment at its base to the tip with a small
ruler. The snood was measured again after being stretched by attaching a clip to the tip of
the snood and pulling on the clip with a Pesola scale to a tension of 30 g. The maximum
vertical and horizontal diameter of each frontal caruncle, as well as its thickness, was
measured with a ruler. The radii of the frontal caruncles was converted to an approximation
of total caruncle area using the equation for the area of an ellipse (A=abrc). The red, polyp-
like projections on the neck called side caruncles were counted. The width of each half of
the thickened, whitish skullcap was measured by placing the ruler in a line from the most
posterior point at which the skullcap halves meet along the middle of the cranium to a point
dorsal to the middle of the eye. The beard was measured from where the "hairs" leave the
skin to its greatest length. Briefly, tarsal spur length was measured from where the spur
enters the scaled skin to the distal tip (Kelly 1975). Tarsus length was measured from the
articulation of the tarsometatarsus with the tibiotarsus to the third scute of the central
phalange of the foot Weight divided by tarsus length was used as an index of male body
condition and is referred to as such or merely as male condition.
Females . Twenty-three, 20-month old female wild turkeys were exposed to an
artificial photoperiod of 14 hours for four weeks prior to, and during, the mate choice tests
in January-March 1993.
Experimental design . The large cage used for housing the females was subdivided
into three sections to serve as a mate choice arena (Figure 3-2a). The female flock was
housed in one half of the aviary during the experiment Each female was admitted singly to
the choice area through a door in the opaque plastic barrier that prevented the untested
females from seeing the displays of the males or the choices made by subject females. An
additional opaque barrier prevented the males from directly interacting with one another.
The subject female was separated from the males by hardware cloth that had been painted
black to make it less visible. Each female was tested only once and was presented with a
unique pairing of males that was never presented to any other female. In total 1 1 males
were used 2-4 times each (mean= 3.3); there was not a significant correlation between the
number of times each male was presented and the frequency of female choice (Spearman
Rank Correlation, r s =0.51, P= 0.14). Males were assigned to the two display areas
randomly on the afternoon before the trial. Only one trial was conducted in any 24 hr
period from 14 Jan-12 March 1993, between 08:00-16:00 hrs. During the ten minute pre-
trial period before the female was admitted, the frequency of spontaneous strutting by each
male and the degree of snood distention (on a scale of 0-4) were recorded. Snood
distention was recorded again immediately after the female was admitted and strutting
frequency was recorded until the female solicited one of the males. Females revealed their
choice of mate by "crouching" (Healy 1992a), a conspicuous mating solicitation posture
exhibited in front of the chosen male. The time spent by the female in front of each male
was also recorded until solicitation. Trials in which females did not solicit within 30 min
were excluded from analysis. The two males from a failed trial were tested again with a new
female the next day.
Live Male 1
Live Male 2
Male Model 1
Male Model 2
Figure 3-2. Mate choice arenas. A) Female is admitted from a holding
area to the choice area where she has a choice of two displaying males; B)
A female is given the choice of soliciting one of two artificial males that
differ only in their snood length and side caruncle number. Screen divider
between males and female is represented by thin line.
Data analysis . The characters correlated with female mate choice were determined
by comparing the characteristics of the male solicited with those of the male not solicited in
each trial. Wilcoxon matched pairs signed-rank tests were used with characters that were
independent of other characters (Siegel, 1956). However some of the measurements of
male ornamentation covaried strongly with each other and with tarsus length (Table 3-1),
making it difficult to tease apart the relative contribution of these characters to mate choice.
Principal component factor analysis was used to isolate independent axes of variation in the
correlation matrix of the five most highly covarying characters (tarsus length, spur length,
skull cap width, stretched snood length and frontal caruncle area). Frontal caruncle area was
transformed by log (x) °- 5 when combined with linear measurements in the factor analysis.
Subsequently these axes were transformed (i.e. rotated) using the varimax solution to
generate four orthogonal axes on which the variables were loaded as uniquely as possible
(Table 3-2). These axes provided four sets of independent axis scores, each set composed
largely of only one character unlike the strongly covarying raw data, that could be used to
determine which of these five characters was most important to a female's choice of males.
For each trial the differences between the scores of the males presented to the female
were calculated by subtracting the score of the male not chosen from that of the male that
was chosen. If females chose males at random with respect to these characters, the
distribution of the differences between preferred and not preferred males would be expected
to have a mean of zero. A positive mean is expected if females are choosing males with the
larger or greater forms of these characters as hypothesized. Because females were used
only once, male pairings were unique in every trial and the difference between male
characters was used in the analyses. This means that each trial was independent despite the
fact that most males were used more than once. Data analysis utilized the "Spin" abilities of
SASjmp (SAS Institute Inc. 1989) and the general statistics of Statview (Abacus Concepts
o >o i— i «— i
^H © OP
rn o O O O .52
o O r- en cs r~
*-«' o ©' ©' o ©'
© m >n tj- c\i >n o>
o oo i— i co i— i "* cs
1-H* ©©'©©©■© ©
t— ( 1/3
^H O ©'©'©©'© © ©
© on so «-« vo co © Tf © © >-;
<-<©'©'©©' ©'©'©'©' p"©'
U OS !_
t- c o o "
13 9 *->
Inc. 1986). Where appropriate, mean values in the results are followed by the standard
error of those measurements.
Male model experiment
Male models . Each female was admitted singly to an arena (Figure 3-2b) in which
she had a choice of interacting with either of two artificial males modified from plastic
hunting decoys of hens. These models were altered in the following ways to make them
appear like strutting males. The heads of the two decoys were painted simultaneously with
alternating brush strokes of enamel paint so that they resembled one another as closely as
possible. The skullcap was painted a light blue, the orbital area a dark blue, the dewlap and
throat were painted red, the sides and back of the neck were painted pink with narrow red
rings applied around the circumference. Natural beards were removed from two domestic
turkey gobblers. Each beard was divided in two, with equal numbers of "hairs," and then
these were trimmed to 17 cm in length and applied to the models.
The decoys have a closed tail, typical of a resting bird. However strutting males
have a fanned tail. Two artificial fans were constructed from colored poster board, glue and
black permanent marker. The vertical diameter of the tail was 33 cm and the horizontal
radius was 69 cm. The fanned tails were positioned just behind the wing tips of the decoys
and held in place from behind with two thin wooden dowels that were stuck in the ground to
stabilize the tail. The models themselves were mounted on a wooden dowel so that their
backs were at a height of 0.65 m. Strutting turkeys drag the primary feathers of their wings
on the ground, but I made no attempt to simulate this posture; the models had their wings in
the normal resting position on their sides.
To provide the sound of strutting, a recording of a single strut from a commercially
available video (Griffen Productions 1990) was played back alternately between the males
so that each male model "strutted" four times per minute. One small 5 watt speaker was
placed to the side and behind each male and aimed towards the area immediately in front of
the male. Using the individual volume controls on the speakers and a Tandy sound meter, I
was able to ensure that the amplitude of the struts were equal (73 db) at a point 0.75 m
immediately in front of each male.
The male models were identical in coloration, tail size, beard length and display
frequency. To test whether the most variable types of fleshy head ornamentation seen in
live males affect mate choice, it was necessary to construct artificial snoods and caruncles
for the artificial males. Four snoods (two 4.0 cm long, two 6.6 cm long), 14 side caruncles
(1.2 cm x 1.9 cm x 1 cm) and four frontal caruncles (2 cm in diameter) were constructed
from latex caulking. The snoods were painted as above so that they were bluish pink, the
caruncles of both types were painted red. These were applied with small pieces of double-
sided tape so that one male model was "more ornamented", i.e. given a long snood coupled
with 5 caruncles on each side of his neck, while the "less ornamented" male was given the
shorter snood and only two caruncles on each side of his neck. Both males had two frontal
caruncles placed side by side at the base of the front of their necks. Only snood length and
side caruncle number were varied because these characters were the most variable in wild-
caught males. One or more of the bodies, tails, beards, snoods or caruncles were exchanged
between the models after each trial and the position of the more ornamented male was
randomized. Thus female preference could be attributed to the only consistent difference
between the males: snood length and side caruncle number.
Females . Twenty-three 17-month old female wild turkeys were implanted with a 21
day-release pellet of 25 mg of estradiol (Innovative Research of America, Toledo, OH,
USA) on 3 Dec. 1992. After 10 days, when most females were exhibiting sexual behavior,
mate choice trials were conducted.
Experimental design. Male model trials were conducted between 08:00-16:00 h
from 13 Dec-15 Dec 1992. The playback of the strut recording was started before each
female was admitted singly to the arena via a remotely opened door. Starting the moment
the female stepped into the middle of the arena from the waiting area, I recorded the amount
of time she spent in each model's half of the arena. A trial ended when the female solicited
or after 20 min had passed. Each female was tested only once.
Data analysis . If females chose mates at random with respect to male snood length
and caruncle number, an equal number of more ornamented and less ornamented males
should be solicited. However if females prefer males with the more elaborate form of these
traits, the more ornamented male should be solicited significantly more often. A Binomial
test (Siegel 1956) was used to test the statistical significance of the results.
Correlates of Ornamentation in Wild Males
Free-living male wild turkeys were captured and measured to determine if male
ornamentation serves as a good indicator of male viability. Yearling male wild turkeys were
captured with bait drugged with alpha chloralose (Williams & Austin 1988) at Paynes
Prairie State Preserve in Alachua Co., Florida, USA between February and March of 1991
and 1992. Ornamentation was measured as described above for captive males, with the
exception of skullcap width, which was not measured. Drugged birds were kept indoors in
boxes specially designed for transporting wild turkeys until they had recovered sufficiently
for safe release (24-96 hr).
The number of attached ticks on the head and neck and on the undersides of both
wings were counted. In addition the lice on the feathers of the chest, back and rump were
counted by lifting and scanning the feathers of each region for two minutes. Blood smears
were made from each individual using blood collected with a heparinized capillary tube from
the alar vein. The smears were air dried, fixed in 100% methanol, and stained. Blood
parasite burden was measured as the number of blood cells infected with hematozoa seen
per 30 minutes of scanning under oil immersion at lOOx. Fecal samples were collected 3x a
day from the boxes of the recovering birds. One gram from each sample was mixed with
20 ml of a saturated NaN03 solution and filtered through cheese cloth into a centrifuge
tube. The tube was filled to the very top so that a coverslip placed on the tube opening
would remain adhered there during the subsequent 5 minutes of centrifugation at 1500 rpm.
During centrifugation most parasite "eggs" either float to the top and adhere to the coverslip
or sink to the bottom, depending on their specific gravity. After centrifugation the coverslip
was placed on a microscope slide and thoroughly scanned. Parasite eggs were identified
and counted directly or when present in very large numbers, estimated with the use of a
McMaster's slide. There are at least 6 species of eimerian coccidia that infect turkeys
(Davidson & Wentworth 1992). Unfortunately, these species are difficult to identify and
could only be categorized as having large or small oocysts in this study. The sediment of
each centrifuge tube was throughly examined for trematode and other eggs under a
Additional measures of ornamentation and blood smears were collected from eight
hunter-killed wild turkeys at Camp Blanding Wildlife Management Area, Florida, in March
Live Males Experiment
Eighteen of the 23 females tested signalled their choice of mates by soliciting
copulation from one of the two males presented in each trial. Traits of males solicited and
not solicited are compared in Table 3-3. Females strongly preferred males with
comparatively higher scores on the second rotated component, i.e. those with longer snoods
and wider skullcaps (unpaired t-test, one-tailed, v =17, mean=1.01, t=3.65, P=0.001). The
length of the male's snood during display, the character state actually visible for assessment
by females, was strongly correlated with the relaxed and stretched snood measurements
(r s =0.80, n=l 1, P=0.01; r s =0.75, n=l 1, P=0.02, respectively). The other variables that
loaded on the first, third and fourth axes: tarsus length, frontal caruncle area and spur length,
respectively, did not account for female choice of males (t-tests, one-tailed, P>0.05).
Characters analyzed individually, i.e. pre-trial strut rate, trial strut rate,
»n in m
CS — i
o o o o
co *— I co co
q « M q
GO G G
O G G O
S ii ±1 ±1 ±L ±1 ±1 3 it it il
en cs in t-; cs p t*; cs i cs -*t | *H
no «s| "^ ts "tf rf r-^ t-4 Tf" od i-I
<-j p >n cs i cs | ^t ; *>o p cs >n ©
©' o ^' Q Q eo <— I d © O ©
+1 -ft +1 +1 +\ +1 +1 +1 +1 +1 +1
\^ N^^ ^^ >w' ^-^ >w' >*-^ >— ' ^S 'W ^-^
in i-j i vq co cs © r~; cn O) oo cs
\d cs" cn cn in «-5 © i-h -<fr o\ »-J
^h ^h cn cs <— i
cn <— i cn /— \
CS CS CS OO
>n cs »-h r-
— i in cs ©
T 1 "^ a o «
=3 S S £.g
oaf S o «
CS v© f-H CS ^
cs_ © -^ ; •"* «n
cs' cn © © P
+1 +1 +1 +1
<-J Tf CS c^
cn cn © ^h
CO p 2
male condition, frontal caruncle depth, side caruncle number, and beard length, also did not
explain female mating preference (Wilcoxon signed rank tests, P»0.05). In the live male
trials females spent significantly more time with the male that they solicited (1.7 ± 0.72 min
for the chosen male vs. 0.42 ± 0.17 min for the male not chosen; Mann- Whitney U-test,
ni=17, n2=17, U=74, U ! =215, P=0.02). The mean time until solicitation was 2.1 ± 0.85
min. Time until solicitation was not dependent on the similarity of the scores or measures
of any of the variables of the two males (Spearman rank correlation, P>0.05). The three
females who did not choose the male with the higher of the two scores for the second factor
seemed to mate more quickly on average than females who did (1.0±0.8 min vs. 2.3±1.0
Male Models Experiment
Of the 23 females tested in the male model trials, only nine solicited copulation from
the decoys. One female solicited immediately in the center of the entrance way of the arena
and thus her choice was ambiguous and this trial was excluded. Of the eight remaining
females, who clearly solicited on only one side of the arena, seven solicited before the more
ornamented model. This preference for the more ornamented model was statistically
significant (Binomial Test, one-tailed, P=0.035). Six of the eight females spent more time
with the model male that they subsequently solicited; however, this difference is not
statistically significant (Binomial Test, one-tailed, P=0.145). The two females who did not
spend more time with their preferred male also took much longer to choose (mean= 9 min
vs. 1.7 min, overall mean= 4 min). Three of the females that chose the more ornamented
model never spent any time on the side of the less ornamented decoy. The single female
that solicited the less ornamented decoy never visited the more ornamented model but
solicited in a typical amount of time (1.88 min).
Nine of the 14 nonsoliciting females were obviously distressed, as evidenced by
frequent alarm calling or frantic attempts to exit the arena, despite a training period. These
females showed no preference for either side of the arena (Binomial Test, two-tailed,
P=0.5). The remaining five females who neither solicited nor alarm called engaged in
nonsexual behaviors such as dust-bathing or preening. They also showed no preference for
either half of the arena (Binomial Test, two-tailed, P=l).
Correlates of Ornamentation in Wild Males
Nine yearling male wild turkeys were captured, measured and screened for parasites
(Table 3-4). All males sampled for these ectoparasites (7 of 9) had attached ticks
(Argasidae and Ixodidae). Ticks were most common on bare areas along the alar vein under
the wing, but also occurred on the back of the head. The total tick burden was not
significantly correlated with any measure of male ornamentation or body size (all P >0.05).
The number of ticks attached on the head, however, was positively correlated with the
relaxed length of the snood (r s = 0.7, n= 9, P=0.05). Lice were most common on the rump
(only one of seven males had no rump lice) but were not correlated with male ornamentation
(all P> 0.05). Half of the eight males sampled for back lice were infected. Numbers of lice
on the back were significantly negatively correlated with the number of side caruncles on
males (r s =-0.66, n=8, P=0.03) and showed a similar trend with the length of the tarsal spurs
(r s =-0.53, n=8, P=0.07). Total lice numbers were not correlated with any measure of male
ornamentation, size or condition (all P»0.05).
One to five fecal samples were collected from each bird (mean=3.4±0.6). Fecal
analysis revealed two parasites to be common: a protozoan coccidia (Eimeria spp.) and a
cecal nematode (Trichostrongvlus tenuis) . Eggs from an unidentified cestode were also
detected in one fecal sample flotation. Sediments from the fecal samples appeared
completely free of parasite eggs.
All nine males had infections with the large coccidia and eight of nine were infected
with small coccidia. Eimeria burdens were not negatively correlated with tarsus length, mass
or condition (all P» 0.05). However coccidia burden was negatively correlated with some
measures of male ornamentation. Average burdens of the large eimerian were
Table 3-4. Parasite loads of one-year-old, wild caught, male wild turkeys.
Parasite % Occurrence (N) Mean Load (±SE)
head 100 (9) 8.4 (±3.8)
wing 100 (9) 17.3 (±3.5)
total 100 (9) 25.8 (±5.9)
back 50 (8) 2.0 (±1.0)
rump 86 (7) 6.8 (±2.3)
total 100 (7) 7.0 (±2.8)
large 100 (9) 342.2 (±190)
small 89 (9) 2276.2 (±2042)
total 100 (9) 2618.4 (±2214)
total 56 (9) 22.3 (±14.81)
Haemoproteus 86 (7) 7.3 (±2.5)
Leucocytozoon 14 (7) 0.1 (±0.1)
Plasmodium 43 (7) 0.6 (±0.3)
significantly negatively correlated with the relaxed snood length of males (r s =-0.78, n=9,
P=0.03). Average burdens of the small eimerians showed a similar, though nonsignificant,
negative trend with relaxed snood length (r s =-0.63, n=9, P=0.07). The overall average
coccidia burden showed a very strong significantly negative correlation with relaxed snood
length (r s =-0.84, n=9, P=0.02).
Individuals with extremely high burdens of coccidia in one sample (e.g., tens of
thousands) never had samples with very low numbers of oocysts. Individuals with lower
burdens (e.g., hundreds of oocysts) generally had at least one sample in which no or few
oocysts were detected. It appears that heavily infected animals are subject to a more chronic
and potentially more debilitating release of oocysts from the gut than are animals with lower
infections, perhaps because more species of coccidia are causing the infection. If this is the
case, perhaps the minimum sample is a better indicator of susceptibility. Relaxed snood
length was not more strongly correlated with the lowest (or highest) burden sampled than
with the average total burden, although these measures were significantly or nearly
significandy negatively correlated with relaxed snood length (r s = -0.55 to -0.82, n=9, 0.02
<P> 0.09). Stretched snood length on the other hand was not significandy associated with
average coccidia burden but was significantly negatively correlated with the lowest
measurements of the large, small and total eimerian burdens (r s = -0.66, -0.70, -0.70,
respectively; n= 9, P< 0.05). Also there were nonsignificant negative trends in correlation
between large, small and total lowest coccidia burden per male and frontal caruncle area (all
r s = -0.63, n=9, P=0.06) and between lowest burdens of small coccidia and frontal caruncle
depth (r s =-0.74, n=8, P=0.06). The tendency of some ornaments to be correlated with the
minimum sample burden suggest that this may be a more sensitive measure of parasite
burden than average or maximum burdens.
The thread nematode fTrichostrongvlus tenuis ) was present in five of the nine males
sampled. Average nematode egg burdens were not correlated with male ornamentation,
body size or condition.
Data from hunter-killed animals showed that relaxed (Mann- Whitney U test, U=0,
U!=16, ni=4, n2=4, P=0.03) or stretched (Mann-Whimey U test, U=0, U!=12, ni=4, n2=4,
P=0.03) snood length could be used to distinguish between first year and older males.
Over the three year classes measured, relaxed snood length increased significantly (rho=
0.79, n=8, P= 0.04) but stretched length only showed a nonsignificant trend (rho= 0.71,
n=8, P= 0.09) with age. These correlations became nonsignificant when the effects of body
condition were removed by partial regression. However relaxed snood length remained
strongly correlated with male condition if the effects of age are removed (rho=0.56, n=13,
The results provide support for a parasite-driven explanation for the maintenance of
female choice for male ornamentation in wild turkeys (Hamilton & Zuk 1982). They do not
support the arbitrary preference model. Females preferred to mate with males that had
longer snoods and broader skullcaps; two tightly covarying male traits. Although skullcaps
were not measured on wild individuals, stretched and relaxed snood length were longest in
individuals with lower and less chronic burdens of coccidian oocysts. Surveys in the wild
have shown this parasite to be quite widespread; 46-100% of turkey poults are infected and
40% of droppings collected from adults contain oocysts (reviewed by Davidson &
Wentworth 1992). Perhaps because this parasite is often ignored during searches for
internal parasites (Ruff et al. 1988), coccidiosis is not commonly suspected to contribute to
the mortality of adult wild turkeys. However laboratory transmission of Eimeria from wild
individuals to domestic poults proves these parasites have the potential to be highly
pathogenic (Prestwood et al. 1973). Chronic effects of coccidiosis in domestic poultry
include delayed maturation due to decreased plasma testosterone levels (Ruff 1988),
decreased egg and sperm production and lower fertility (Bressler et al. 1951; Ruff et al.
1984; Ogbuokiri & Edgar 1986; Ruff & Wilkins 1987). If similar effects occur in wild
populations, and if resistance to coccidia in wild turkeys is heritable as it is in domestic
poultry (Johnson & Edgar 1982), females that choose mates that are resistant to Eimeria
would have greater fitness because their offspring are relieved of the deleterious effects of
these parasites. Thus female choice for male snood length appears to be a behavioral
adaptation against parasitism.
The beard, frontal caruncles and other aspects of male ornamentation were not
subject to female choice, despite the fact that there is some indication that they may also
indicate parasite burden. Why did females not discriminate among males by these
characters? There are three possible explanations. First and most obvious, is that the captive
turkeys of the eastern subspecies are not subject to the same selective pressures as the wild
osceola males sampled in Florida. Limited data suggest that parasite burdens may be lower
in the northern parts of the wild turkey's range (Sasseville et al. 1988) Second the apparent
trends between caruncle quality and other parasites may be the spurious result of using
multiple statistical tests. Larger sample sizes of wild males are needed to document the
reliability of these other ornaments as indicators of parasite burden to females. Last, studies
of female choice in other galliformes have shown year to year shifting in some of the male
characters that are assessed (e.g., Zuk et al. 1992), suggesting that frontal caruncles, which
appear to be inversely correlated with parasite burden, may be important in mate assessment
in other years.
Threadworms, Trichostrongylus. have a dramatic effect on the fitness of some other
galliforms, particularly grouse (Hudson & Dobson 1991), but infections in wild turkeys are
generally not severe enough to cause pathologies (Davidson & Wentworth 1992) and
therefore it was not surprising that these nematodes did not affect the degree of male
ornamentation. Similarly ectoparasite burdens were low and probably have a subde,
cumulative effect on male condition not readily revealed by ornamentation (e.g., Booth et al.
Because harsh winters limit survivorship in wild turkeys, it was hypothesized that
females should try to assess male ability to survive such conditions. It was hypothesized
that turkey hens could assess male condition by the male's ability to perform energetically
costly displays. This hypothesis was not supported directly. Male condition and display
frequency during the mate choice trial did not explain variation in female choice. In fact
some females solicited so quickly that they would have had little opportunity to measure
male strut frequency. However the characters females assessed during mate choice, snood
length and skullcap width, were positively correlated with the male's average strut frequency
over all his trials (Table 2-1). In the well-fed captive turkeys neither average strut frequency,
relaxed or stretched snood lengths nor display snood length were correlated with male
condition. But after statistically removing any effects of age, the stretched snood length of
wild males was significantly positively correlated with male condition and age was no
longer correlated with snood length. Although male strut frequency was apparently not
assessed by females during the mate choice trials, females mated with males that strutted at a
higher rate and that were in better condition by choosing males with longer snoods and
wider skullcaps. Further field captures and experimental manipulations will be necessary to
determine if display rate is a specific indicator of the stored fat resources of the male.
Male age is perhaps the best indicator of a male's good genes. Older males are
probably not a random sample in terms of natural selection; they have survived multiple
challenges over time, including parasitism, starvation and predation. One hypothesis for the
maintenance of male ornamentation in wild turkeys is that these characters serve as
indicators of male age. In this study captive females did not choose among 2 year old males
based on previously documented age class indicators, such as spur length. This may be
because these cues provide no information when males are very similar, i.e. within one age
class. Unfortunately previous studies of correlates of wild turkey age have ignored the head
ornamentation of males (Kelly 1975; Steffen et al. 1990). Results from a limited sample of
wild males covering three age classes shows that the characters females are assessing in
mate choice are not good indicators of male age because age covaries with body condition.
A larger sample size is needed to tease apart these two variables. For example, more field
data are needed to determine if the snood lengths of heavily parasitized, older males are
shorter than lightly parasitized, younger males. In any case snood length within one age
class appears to be a reliable cue to the body condition of the bearer as well as to the
coccidia burden among males.
The handicap model for the maintenance of male ornamentation proposes that
females assess a male's ability to survive despite his burdensome ornaments. Field
experiments that record the reproductive success and survivorship of manipulated males are
necessary to test this hypothesis. Nevertheless anecdotal observations provide some
support for the handicap model. The conspicuous ornamentation and courtship displays of
male wild turkeys attract predators and are thought to increase their risk of predation (Miller
& Leopold 1992). The uninsulated bare head of males may also represent a
thermoregulatory handicap. Birds generally lose large quantities of heat from feathered
heads (Hill et al. 1980). Unfeathered heads such as the turkey's should lose even more heat
and may hasten male starvation during severe winter storms. Since there is no clear
advantage to having bare heads in winter, it seems reasonable to suggest that the bare heads
of male wild turkeys may have evolved as a handicap. Similarly other forms of male
ornamentation, such as the beard and fanned tail, might serve as handicaps.
This study suggests that the snood and skullcap of males are maintained by female
choice for males with low burdens of coccidia. Also these traits may be used by females to
choose older males as mates and perhaps males that have more body fat Snood length and
skullcap appear to be the only aspects of male ornamentation under direct selection by
females. The selective pressures maintaining the striking array of extravagant
ornamentation of wild turkeys is only partly explained. Males have strikingly colored
caruncles on their head and neck, a long beard projecting from their chest, iridescent
feathers and a large fan-like tail that are apparently not subject to female choice, but may
instead play a role in male-male competition (Chapter 4) or be maintained because of its
other functions (Chapter 5).
MALE DOMINANCE AND VARIATION IN
FLESHY HEAD ORNAMENTATION IN WILD TURKEYS
The mating system of wild turkeys has been classified as male-dominance polygyny
(Williams & Austin 1988). In the breeding season males vocalize ("gobble") to attract
flocks of receptive females and then follow and display to those hens throughout the day.
During this time other males may be attracted to the same females and battles between males
over mating access often result (Watts & Stokes 1971). Because males are unable to obtain
exclusive access to females by defending food or other resources that female turkeys need
(Emlen & Oring 1977; Hurst 1992), they control access to the females themselves. As a
result male wild turkeys engage in frequent, direct, agonistic interactions that determine
mating success (Watts & Stokes 1971). These agonistic interactions begin when the males
are juveniles and establish dominance heirarchies within their age cohort (Healy 1992a) and
continue as males encounter and fight with other males after dispersal from the natal
territory and in subsequent breeding seasons (Watts & Stokes 1971; Williams & Austin
1988; Healy 1992a).
The War Propaganda model (Fisher 1958; Borgia 1979) proposes that in mating
systems such as the wild turkey's, in which females gain no direct benefits from males and
males compete for access to females, male fighting success should be assessed by females
during mate choice. Females should do this because fighting ability is probably a good
measure of overall male condition. Another reason to mate with dominant males is that the
components of dominance may be heritable (Dewsbury 1990); therefore, a female that
mates with a dominant male may have higher fitness because of an increased likelihood that
her mature male offspring will be able to gain access to hens. But females are not the only
sex in which selection should favor the ability to recognize dominant males. To avoid
battles in which the risk of injury outweighs gains in lifetime fitness, males should assess
the condition of their competitors before fighting. Both the receiver and the signaller might
benefit from the transmission of information about male condition (Rohwer 1975; Rohwer
1977). Dominant males avoid the costs of continually battling weaker males for preferential
access to resources and subordinate males avoid the risk of serious injury from dominant
males who are likely to be healthier and stronger and win fights (M0ller 1987). Thus both
females and males should assess the dominance status of males.
Fleshy ornaments are associated with intrasexual competition in many galliform
species (Stokkan 1979; Brodsky 1988; Ligon et al. 1990; Holder & Montgomerie 1993)
and are uniquely suited to signal male status honestly. These structures generally develop
only upon sexual maturity and their size in males is known to be testosterone dependent in
some species (Allee et al. 1939; Collias 1943; Stokkan 1979). Higher testosterone levels
are often associated with dominance and increased aggressiveness in birds (Wingfield et al.
1987), suggesting that fleshy ornamentation may correlate with a male's dominance status.
Dominant males might have relatively higher testosterone levels because they are in better
physical condition (Ligon et al. 1990) or because they are able to suppress the testosterone
levels of subordinate males (Lisano & Kennamer 1977).
As demonstrated earlier (Chapter 3), the snood length of male wild turkeys is a
indicator of coccidia burden, condition and possibly age. This would suggest that males
could use this aspect of male ornamentation to assess the vigor of males they encounter and
thus consider the risks of combatting those males. If this is the case, males with longer
snoods would be expected to be dominant over less ornamented males, and that males with
similarly sized ornamentation should be in the most conflict over relative status. In this
study I test this hypothesis by comparing the ornamentation of males matched in dyadic
dominance trials. However males may establish dominance not because of ornamentation
itself, but because of strongly correlated attributes such as body weight Therefore I also
examine the response of captive males to the presentation of artificial males that differ only
in snood length and caruncle number.
A detailed description of the rearing of the wild turkeys used in these experiments,
including cage sizes is provided in the previous chapter (Chapter 3). Briefly, the wild
turkeys were obtained from a game farm as day-old chicks. At four months of age the
females were separated from the males and the males were divided into two, visually isolated
groups. Pilot observations during the spring 1992 breeding season showed that very few of
the males actively strutted when group housed. Therefore when the males were 15 months
old, six months before the 1994 season, the males were moved to individual pens separated
by opaque dividers. This had the effect of stimulating more males to perform the strut
display. Eleven of the 28 males displayed spontaneously and regularly in their individual
pens. These animals were designated "displayers" and were used in the mate choice trials
(described in Chapter 3) before being tested in the dominance trials. For the dominance
trials, conducted in Feb-April 1993, a male from one rearing flock was matched with a male
from the other flock. These males were chosen from their respective flocks at random
except that "displayers" were always matched with other displayers. An additional male
who displayed occassionally was used in the dyad with the odd numbered "displayer". The
pairing of males based on display frequency mimics the situation in the wild where
displayers usually fight one another and are avoided by nondisplaying individuals and
juveniles (personal observation). Each dyad of males was measured for ornamentation and
body size, as described previously (Chapter 3), on the day before the dominance trial.
In the live male dominance trials, two males were admitted to a central arena (2.7m x
5.3 x 2.7m) from their respective holding pens (Figure 4- la). The males were familiarized
with the arena for several hours on the day prior to the trial and had spent part of their
rearing period in the pen as well. No food or mates were provided in the arena, although the
males often found something to peck at on the dirt floor of the pen. This situation
simulated a random meeting of unfamiliar males on common ground with no resource to
contest. Trials lasted 10 minutes and were videotaped so that the rapidly occurring
dominance interactions could be reviewed and transcribed later.
A heirarchy of criteria was used for designating one member of the dyad as
subordinate and the other dominant Most males actually fought with one another.
Fighting begins when one male approaches the other and begins giving "fighting purrs",
vocalizations that elicit combat (Healy 1992a). The other male typically responds with
fighting purrs until one of them slaps the other with a wing while simultaneously kicking
him. The birds continue to wing slap and kick one another until one of them suddenly turns
and attempts to flee. Fleeing males are generally pursued and sometimes pecked on the
head. If the males fought, the fleeing male was designated the subordinate. In some trials
one male fled immediately upon encountering the first male and as a result was designated
the subordinate. In one trial the males did not fight nor did one flee from the other. In this
case the male that was displaced while pecking at the ground was designated the
For each trial I recorded the number of wingflapping bouts given by each bird prior
to physical contact (pecking in the absence of fights, or wing slapping/kicking in birds that
fought) and latency to physical contact. A wingflapping bout consisted of the male standing
in place and flapping the wings two to five times. This action pattern is often used in
aggressive contexts in Galliformes (Kruijt 1964; Buchholz 1989). If the males fought I
recorded the identity of the male who began giving "fighting purrs", the frequency of
Figure 4-1. Male dominance trial arenas. A) Randomly matched males
are admitted to a central arena from holding pens; B) A living male is
given the choice of feeding from in front of two artificial males of
different snood lengths and side caruncle numbers. Another male waits in
a holding area.
pecking and kicking by each male during combat, the duration of the fight and the identity
of the male that stopped fighting first and began to flee. A dark shelter was provided in the
rear of the arena so that the subordinate male could hide from the aggressor. None of the
males was injured.
The characteristics of the dominant male were compared with those of the
subordinate in each trial with one-tailed Wilcoxon matched pairs signed ranks tests (Siegel
1956) or one-tailed, unpaired t-tests on principal component factor score differences. The
differences between the factor scores of the two males used in each trial, calculated as the
dominant male's scores minus the subordinate male's scores, are expected to be significantly
greater than zero if dominant males are more highly ornamented (as described in the
methods for Chapter 3). Correlations between time measurements and differences in male
ornamentation were determined using the Spearman rank correlation coefficient
Differences in ornamentation between subgroups of males, e.g., displayers and
nondisplayers, were tested for statistical significance with the Mann-Whitney U test.
Male Model Trials
Artificial males constructed from decoys were used to test the effects of male fleshy
ornamentation on the decision-making of live males. The use of models allowed for the
experimental control of male characteristics that usually covary with snood length. The
model males were exactly like those used in the mate choice experiments described earlier
(Chapter 3), except that they did not have fanned tails and they were not accompanied by the
recorded display sounds. One male model had a long snood and many side caruncles
(highly ornamented) and the other had a shorter snood and fewer side caruncles (slightly
ornamented). The models were placed 1.3 m apart at one end of the rectangular arena
(Figure 4-lb). The relative positions of the highly ornamented and slightly ornamented
models were randomized. Approximately 15 ml of bird seed was placed in a pile on the
ground immediately below the bill of each model. Males were denied food for a few hours
immediately before the trial, but this appeared to have little impact on their behavior because
they did not feed immediately when food was returned after the trial. In any case bird seed
was a preferred treat that the turkeys would eat preferentially even when their regular food
was available ai libitum . During each model trial a single male was admitted through a door
in one side of the arena. The latency of the males to feed from one of the seed piles was
recorded. Males that did not choose a seed pile within 30 min were removed and were not
Live Male Trials
Fighting occurred in 8 of the 14 trials. Fights lasted on average only 46.1 sec. Out
of the 6 trials in which no fighting occurred, there were three cases in which the subordinate
male fled from the other male immediately. In one case the males did not fight nor flee, nor
did they ever direcdy contact one another. The duration of fighting showed a trend toward
being negatively correlated with differences in beard length between the two males (r s =
-0.51, p=0.07). During fighting only 2 of the 28 males tested, distended their snoods
slightly (just past the bill opening). In both cases these males were later classified as
dominant to the individual with which they were paired.
Seven of the ten male characteristics measured directly were found to be strongly
correlated (r 2 >0.4) with one another (Table 4-1). These were spur length, relaxed snood
length, side caruncle number, skullcap width, frontal caruncle area, frontal caruncle depth
and body mass. A factor analysis utilizing varimax transformation produced four factors
from these variables (Table 4-2). Snood length and mass loaded on the first factor, spur
length, frontal caruncle area and caruncle depth loaded on the second factor, skullcap width
loaded on the third factor and side caruncle number loaded on the fourth factor. The factor
scores of these variables are analyzed below in addition to the raw measurements in an
attempt to identify underlying axes of variation in these characters.
i — t
u c u a H""
J3 «S &
^2 o 5 -
^h ^ _, Tt o o r-
N^N'i 4 s } """* O
o o o o o o o
d © © © © © d
en M cxj ^t cj\ o ^ <?\
r-^q o i; H _ >n Ui «N
d d c: <^> d d d d
The differences between the two males scores on factors 2, 3 and 4 were not
significantly different from zero (mean differences = -0.14, 0.5, 0.11, respectively; Table 4-
3). However the results for factor 1, though only nearly significant, suggest that snood
length and body mass may be associated with dominance in male wild turkeys (mean
difference = 0.48). When the difference between dominant and subordinate males in these
characters are assessed direcdy, without the aid of factor analysis, relaxed snood length is
the only character that is significantly different between the two groups (Table 4-3).
However dominant males on average had higher values for all ten of the variables measured
(Binomial test, N=10, one-tailed, P=0.06).
Individuals that started fights were no more likely to become dominant than those
that responded to the threat behavior (X 2 =2.27, v=l, p> 0.05). The time until first contact in
the 13 trials in which contact occurred averaged 129.2 sec, although 85% were 32 sec or
less. The difference in ornamentation of the males was associated with the time until first
contact between males that actually fought The time until first contact was positively
correlated with differences between males on the second factor (spur length, caruncle area
and depth; r s = 0.52, n=13, p=0.07). This relationship was statistically significant if the raw
differences in spur length are tested rather than the factor scores (r s = 0.59, p=0.04). The
difference in the frequency of wingflapping bouts prior to contact, on the other hand, was
negatively associated with time until first contact, although only nearly significantly (r s =
The males had been divided into two matched groups for the dominance trials:
nondisplayers and displayers. The eleven displayer males tended to have longer stretched
snoods (4.0 vs. 4.6 cm; U=56, U!=131, p=0.08) and larger factor two scores (spur, frontal
caruncle area and depth; U=52, U ! = 135, p=0.05). However only caruncle depth was
significantly different when the raw values of these variables were compared between the
two groups (1.0 vs 1.2 cm; U=30.5, U l = 156.5, p=0.003).
9 01 co o)
© a e c
oa cw os
G G G
i-h r-i os cs co i-j en i-H cs en ©
so ri er i cn Tf ci i> *-! ■<* oo o
t-H i— i tN SO
/* — N r — ^ r ~\ r -*
cs en en cs
rf r~. m so
Tt rH o\ M « VO O fH -h W O
■g cSONt^'— i»0 ,: t v O'-jr-;t~-;<^
a so t-h -<t en ■<* «n cs «-i <tf" od o
T3 5 S U «S
?> © Q »^h r— < »rt
O On sO #
es es es en
rj- OO v> so
CN SO CS O
S n Q
"- 1 -G
en r- ■*
Male Model Trials
Twenty one of the 28 males tested responded by feeding within the 30 min trial
period. Of the seven males that did not respond, two attacked the decoys (one gave fighting
purrs to the less ornamented model, the other pecked the snood of the more ornamented
model), four did not feed or approach the decoys, and one trial was excluded because a
decoy toppled over. A significant number of the 21 males that responded under the trial
conditions fed in front of the less ornamented decoy (X 2 = 8.0, v=l, p< 0.01). The four
males who chose to feed from the seed in front of the more ornamented male tended to have
smaller spurs (mean= 2.1 vs. 1.3 cm; U= 22.5, U*= 73.5, p=0.09) and fewer side caruncles
(25.7 vs. 12.8; U=18, U 1 =78, p=0.05) than the 17 who fed in front of the less ornamented
model. The latency of the males to feed (i.e. the time from the start of the trial until males
began to feed) was not correlated with any of the male characters measured. The males who
did not feed during the trial were not different from the other males in any of the measured
characters (Mann Whitney U-tests, all p> 0.05).
In dyadic encounters between wild turkeys the only male character significantly
predictive of the outcome of male-male interaction was relaxed snood length. Dominant
males had longer relaxed snoods than their subordinate partners. Similarly males tended to
avoid feeding on seeds near model males with longer snoods in the male model trials.
These complementary results strongly suggest that males assess one another's snood length
as an indicator of male condition and status. Therefore it is surprising that differences in
snood length do not affect the time until, or duration of, combat If snood length is a good
indicator of status, males with similar snood lengths should need to battle longer to resolve
relative dominance status than pairs of males with very disparate snood lengths. Additional
studies are necessary to understand why males do not adjust their investment in fighting
according to their likelihood of winning.
Differences in several other male characters showed tendencies to be correlated with
the time until first contact and the duration of fighting. Males with similar spur lengths
fought each other sooner after meeting than dissimilar males. Spurs serve as weapons
during battle in a number of galliforms (Davison 1985) including wild turkeys, but none of
the evidence shows that the length of spurs affects the outcome of combat. In fact an
experimental field study in Sweden clearly demonstrated that variation in the length of spurs
is not a good predictor of dominance status in ringneck pheasants (Phasianus colchicus ).
although this character is assessed by females during mate choice (von Schantz et al. 1989).
Spur length in wild turkeys is a reliable predictor of age (Kelly 1975) and older males tend
to be larger and in better condition (Chapter 3). Therefore spur length may normally be
assessed by competitors before battle in the wild where different aged animals encounter
one another, but is of little significance in the outcome of trials where males are of the same
age as was the case here.
Differences in the frequency of wingflapping bouts, a behavior associated with
aggressiveness in other galliforms, tended to be negatively correlated with time until first
contact However this behavior did not occur at a higher frequency in dominant individuals.
Similarly differences in beard length between males tended towards a negative correlation
with duration of fighting. The greater the difference in beard lengths between males the
more quickly they tended to resolve fights. However beard length was not a good predictor
of dominance status either. Beard length increases with age to some degree (Kelly 1975),
but is strongly affected by abrasion and thus is probably not a consistent indicator of age
across different habitats (Pelham & Dickson 1992). Studies of beard and spur
development relative to agonistic interactions in mixed-age populations are needed.
In captivity males classified as "displayers" had longer stretched snoods than
nondisplayers. I did not test the relative dominance of these two types of males. Off hand
it might seem reasonable to conclude that displayers will be dominant over nondisplayers
because displayers have longer stretched snoods. However more caution in interpreting the
implications of stretched snood lengths is necessary. Snood distension is muscular (Lucas
& Stettenheim 1972), therefore the difference in stretched snood length between displayers
and nondisplayers may possibly be attributable to the greater exercise given the snood by
the "displayer" males. Additional evidence to support this hypothesis comes from the
previous chapter (Chapter 3) where average display rate (a measure of how often the snood
is distended) was found to be strongly correlated with stretched snood length but only
weakly correlated with relaxed snood length. For this reason it is unclear whether
displaying males can be assumed to be dominant over nondisplayers based on
ornamentation differences. This cautionary note serves as a reminder that behavior can
influence morphology just as the reverse is true.
Snood length was strongly correlated with body mass, a factor thought to be very
important to resource holding capacity and the outcome of dyadic encounters in many
species (Parker 1974; Richner 1989; Beaugrand et al. 1991). In this study males differed
by several kilograms in weight in some cases, and yet this variable, when measured alone,
did not have a significant effect on the outcome of fighting. Snood length, on the other
hand, was greater in dominant males when raw measurements were used and nearly so when
factor scores were used. This suggests that snood length indicates more than just body
mass to potential competitors. Results presented earlier (Chapter 3) confirm that snoods are
good indicators of a number of measures of male condition, including parasite burden,
energy reserves and possibly age. My results suggest that males use snood length as a
general measure of the risks associated with fighting over food or status and that females
use this male character to assess male condition and fighting success.
Similar results were obtained in comprehensive studies of sexual selection in red
junglefowl, Gallus gaUus : (Ligon et al. 1990; Zuk et al. 1990a; 1990b; 1992). Comb length
in this species is maintained by female choice. Females choose males based on comb
length because this character indicates male parasite burden and fighting success. Thus the
comb of junglefowl and the snood of wild turkeys fit Borgia's (1979) War Propaganda
model for the evolution of extravagant ornamentation because females are choosing males
based on characters that indicate male fighting success.
In contrast studies of sexual behavior and fleshy ornamentation in grouse suggest a
different scenario than that in phasianids. Comb size is a good indicator of male status in
ptarmigans (Gjesdal 1977; Moss et al. 1979; Stokkan 1979) and affects how males interact
(Holder & Montgomerie 1993), but this character is not under direct selection by females.
Instead female ptarmigan preferentially mate with males whose combs show less damage
from fighting. In larger grouse comb characteristics appear to be unimportant to female
mate choice (Alatalo et al. 1991). Why do females assess male fleshy ornamentation
differently in phasianids and grouse? One reason may be that phasianids have permanently
exposed fleshy ornaments, while grouse can hide their combs beneath feathers. The
facultative nature of comb exposure makes it more difficult for females to assess the size of
the comb. As a consequence females are selected to base their mate choice on a more easily
and quickly assessed characteristic (Sullivan 1994), such as scarring and damage from
fighting (Brodsky 1988). This interpretation is at best speculative but warrants further
In conclusion it appears that female and male wild turkeys can use male fleshy
ornamentation to assess male condition and dominance status. Male characteristics that are
important in other species, such as body mass or age, were not valid predictors of the
outcome of male-male agonistic interactions in this study or were not included as a variable.
Fleshy ornamentation appears to be an indicator of male status because it is probably
dependent on a condition- and dominance-dependent hormone, testosterone. Both males
and females seem to take advantage of the information provided by snood length to increase
THE THERMOREGULATORY ROLE OF THE
UNFEATHERED HEAD AND NECK IN MALE WILD TURKEYS
Endotherms usually maintain body temperatures above environmental temperature at
considerable energetic cost An unexplained exception to the general endotherm pattern of
having insulation against heat loss are birds that have brightly colored areas of unfeathered
skin on their heads and necks. Although the bright coloration of these structures is
consistent with a sexually selected function, some studies have suggested that these areas of
bare skin also maintain sublethal brain temperatures by dissipating heat via cephalo-cervical
retes (Crowe & Crowe 1979; Crowe & Withers 1979; LaRochelle et al. 1982; Chapter 2).
The heat dissipation hypothesis is supported by correlative studies showing that unfeathered
head and neck skin is maximally exposed at high temperatures and that in some taxa the
size of unfeathered areas is greater at low latitudes where heat dissipation may be of greater
importance (Crowe 1979; Chapter 2). Highly vascularized fleshy ornamentation presents a
functional puzzle when species are distributed over a large latitudinal range in which they
are exposed to both temperature extremes. Although these species may benefit by using
their FS to dissipate heat under hot conditions, the uninsulated nature of these structures
subjects them to extreme heat loss under cold conditions and heat gain in the presence of
solar radiation. In this study I test the possible thermoregulatory function of unfeathered
head ornamentation in a species that commonly faces extremes of cold and heat, the wild
turkey (Meleagris gallopavo) .
Wild turkeys occur over a broad range of temperature extremes from their southern
limit in southern Mexico to their northern limits along the US-Canada border (Dickson
1992). Males are twice as large as females, and have brightly colored unfeathered skin on
their heads and necks. In addition this skin is covered with polyp-like elaborations of the
integument called caruncles. A thin dewlap extends from the mandible down to the neck.
Perhaps most distinctive is the bare, distensible frontal process or snood that projects from
the forehead at the base of the upper bill.
Fleshy head ornamentation in wild turkeys and other galliformes is often thought to
be maintained by sexual selection, that is, it functions in mate choice and male-male
competition. Ample empirical evidence supports this contention (Brodsky 1988; Boyce
1990; Hillgarth 1990; Ligon et al. 1990; Zuk et al. 1990a; 1990b; Spurrier et al. 1991; Zuk
et al. 1992; Chapters 3 & 4;). A role in sexual selection, however, does not rule out
concurrent functions for these structures in thermoregulation. To understand why the
unfeathered areas of turkeys are maintained despite the possible costs in terms of heat loss,
the thermoregulatory tradeoffs faced by wild turkeys with feathered and unfeathered heads
must be assessed. Because all extant wild turkeys have bare heads, in this study I
experimentally insulate the heads and necks of turkeys, to assess the thermoregulatory
tradeoffs that ancestral turkeys may have faced at cold and hot temperatures.
Subjects and Apparatuses
Eight, two-year old, male wild turkeys, obtained as chicks from a game farm (L&L
Pheasantry, Hegins, PA, USA), were used in the metabolic trials. These birds were reared
as described in Chapter 3. Gray and Price (1988) showed no difference in the metabolic
rates of wild turkeys from game farm or free-living sources. Average body weight of these
individuals was 7.1 kg (range 6.4-8.1 kg). During the study period (June- Sept 1993) the
birds were provided with feed (Purina Gamebird Maintenance, 12% protein) and water ad.
libitum . Subjects were denied food for 26-29 hrs immediately prior to each metabolic trial
to insure that they were post-absorptive. Post-absorptive conditions are necessary to
measure the basal or minimum rate of metabolism. Water was still available during the pre-
Oxygen consumption and total water loss were measured in an open system
(described by McNab 1988). The temperature of the 329 liter metabolic chamber was
regulated by pumping water from a water bath through the chamber's hollow walls. Room
air was sucked through the metabolic chamber, pumped into glass columns filled with soda
lime (to remove CO2) and silica gel (to remove H2O), after which flow rates, which averaged
20.6 1/min, were measured by a Brooks Sho-Rate flowmeter. Subsequently the airstream
was sampled with an Applied Electrochemistry S-3A oxygen analyzer. The temperature
and humidity of room air varied very little, 23.5 °C (±0.0) and 61.2% (±0.2), respectively.
Humidity in the chamber was not controlled. The bird's evaporative water loss was
measured gravimetrically, that is, by weighing the silica gel. Core body temperature was
measured by inserting a copper-constantan thermocouple, tipped with a thin layer of
silicone, into the cloaca of the bird to a depth of 20 cm. This measurement was taken
immediately before the subject was placed in the metabolic chamber and immediately after it
was removed from the chamber. Six surface temperature measurements were taken: feather,
leg, body skin, head skin, frontal caruncle and dewlap. Surface temperatures were measured
with a bare-tipped thermocouple held against the appropriate spot while the subject was still
in its holding box before the trial and again while it was still in the metabolic chamber at the
end of a trial. Skin and feather surface temperature were measured approximately 3 cm
ventral to the carpal joint of the wing. Leg temperature was measured immediately posterior
to the third scale below the tarsal joint on the left or right leg, depending on which was
accessible. Head skin temperature was measured on the back of the head at a point in line
with the lower mandible. Surface temperature of the frontal caruncles and dewlap were
measured at their approximate centers.
Different rates of physical activity across subjects and trials can make it difficult to
detect the effect of experimental treatments on metabolic rate. Therefore I minimized the
bird's activity by conducting trials at night in the dark. Metabolic trials lasting 2.5 hours
were conducted between 20:00-03:00 hours. All subjects were given at least one day
between trials. Individual turkeys were tested at the same time of day (either 20:00 or 0:00
hours) across all treatments to minimize circadian effects on matched comparisons of
metabolic rate. The first 30 min of each trial served as an equilibration period during which
the bird calmed down after handling. The least observed rate of oxygen consumption
(corrected to standard pressure and temperature) measured during each of the four
subsequent 30 min periods was used to calculate an average metabolic rate for the entire
trial. All individuals were given two 2.5 hr habituation trials prior to the experimental trials.
Usually the subjects rested quietly during the experimental trials. Behaviorial states were
recorded by instaneous sampling (Martin & Bateson 1986) every 30 min. Three states
were noted as present or absent: standing, head tucked under feathers, panting.
Observations were made with the aid of a flashlight through a small window in the chamber.
To determine the potential thermoregulatory impact that head feathering would have
on wild turkeys, their thermal balance was ascertained when their heads were "bare" (see
below) and when they were insulated as though they were feathered. To approximate the
insulatory properties of head and neck feathering, the bare head and neck of the turkey
were covered with a double layer (0.6 cm on head, 0.9 cm on neck) of acrylic sock (Adler®
"Casual Acrylic Crew," 75% Hi-Bulk Acrylic/25% Stretch Nylon) with large holes for the
eyes and the entire bill. Any irritational effects of the insulatory head covering on metabolic
rate were controlled by placing a hood made of thin, nylon netting with little insulatory value
on the heads of the "bare" individuals. The control head net and insulatory head socking
were held in place with small alligator clips that attached to the back and chest feathers at the
base of the neck. The efficacy of using head socking to approximate the insulation
provided by normal feathering was determined by studying the warming curves of the
feathered and unfeathered/reinsulated heads of domestic roosters (Gallus gallus) .
Six dependent thermoregulatory variables may be affected by head insulation.
Metabolic rate, as measured by oxygen consumption (cm 3 02/g*hr), is a measure of the
work the animal does to maintain thermeostasis. Rate of evaporative water loss (g/hr) is a
measure of the heat lost via evaporation. Metabolic heat production (Hm) and evaporative
heat loss (H e ) can be converted to common units (mW/g) to compare the cooling capacity
of the animal in different treatments. Cooling capacity is the ability of the animal to
dissipate metabolically produced heat by evaporation. It is expressed as a percentage,
calculated as the heat lost by evaporation divided by the heat produced by metabolism
(100% x He/H m ; Calder & King 1974). Total thermal conductance (mW/cm 2 °C)
measures all the heat lost by the animal, including evaporative heat loss, and is the inverse of
insulation. It is estimated using the values for heat production, and ambient and body
temperatures. Dry thermal conductance on the other hand is a measure of all
nonevaporative means of heat loss: radiation, convection and conduction. If total
conductance is exceeded by heat production, heat is stored in the tissues of the animal and
body temperature rises. Each of these variables may be varied by the animal to cope with
increased head and neck insulation.
Thermoregulatory trials were conducted twice for each turkey at each of three
ambient temperatures (0, 22, 35 °C); one time as a control, the other with its head insulated.
These temperatures were chosen to be below, within and above, respectively, the zone of
thermal neutrality (Gray & Price 1988). The temperatures are also within the range that
turkeys experience in the wild. A total of 48 trials were conducted. This matched design
compares the metabolic values of each turkey in the experimental treatment to the values
obtained from the same bird in the control treatment. This serves to minimize the effects of
inter-individual variation on the effect of the experimental treatment. Due to scheduling
conflicts in the laboratory, every turkey was tested at °C before it was exposed to the other
temperatures. The presentation order of the trials at 22 and 35 °C was randomized.
Repeated measures ANOVAs were used to test the effects of body size (mean = 6.7 vs. 7.5
kg for small and large, respectively), chamber temperature and head insulation on oxygen
consumption (cm 3 02/g*hr), cooling capacity, total and dry thermal conductances (mW/cm 2
°C) and changes in body and surface temperatures (°C). The effect of each 30 min
sampling period was also included when oxygen consumption was the dependent variable.
Treatment groups exhibited similar variances (Fmax tests, all p>0.05; Sokal & Rohlf 1981).
Reported p values have been adjusted using Greenhouse-Geisser epsilon values. This
technique conservatively compensates for the use of repeated measures by adjusting the
degrees of freedom (Abacus Concepts Inc. 1989).
The mass specific rate of oxygen consumption was significantly lower for large
individuals across all temperatures (mean= 0.4140 ±0.007 vs. 0.4730 ±0.013 cm 3 02/g*hr;
Table 5-1). Rate of metabolism was not significantly different for uninsulated and insulated
turkeys at and 22 °C. However at 35 °C insulated turkeys exhibited a significantly higher
average metabolic rate than uninsulated turkeys (Table 5-2, Figure 5-1). A significant,
three-way interaction between head insulation, temperature and time period suggests that the
effects of head insulation became more pronounced the longer the subject was exposed to
the chamber conditions at hot temperatures.
When the cooling capacities of uninsulated and insulated turkeys were compared,
uninsulated turkeys demonstrate a significantly greater ability than insulated turkeys to
dissipate excess metabolic heat by evaporation at 35 °C, but not at cooler temperatures
(Figure 5-2). Total thermal conductance increased with temperature. It was also greater for
insulated birds overall (Table 5-1, Figure 5-3) but this difference was significant only at the
hottest temperature (Table 5-2). Dry thermal conductance decreased with increasing
temperature in the uninsulated birds. Insulated birds showed a similar pattern of
conductances at and 22 °C but had significantly higher values than uninsulated birds at 35
Head insulation resulted in significantly greater core body temperature changes of
birds at 35 °C, but not at the lower temperatures tested (Figure 5-4). Head insulation served
to keep head skin warmer at °C (31.2 ±1.0 °C uninsulated vs. 36.7 ±0.3 °C insulated),
but did not result in significantly higher skin temperatures at 22 and 35 °C. Insulated
turkeys at 22 °C increased their dewlap temperatures significantly more than uninsulated
birds (32. 1 ±0.7 °C uninsulated vs 35.2 ± 0.3 °C insulated), but this was not true at or
35 °C. Body skin, feather, frontal caruncle and leg temperatures all increased with
increasing ambient temperatures, but were not affected by the insulation treatment (Table 5-
Across and within each temperature treatment, head insulation had no effect on the
proportion of instantaneous observations in which the subjects were seen to be standing,
panting, or had their head tucked in back feathers or under the wing (Mann-Whitney U
tests, N=16, all p»0.05). Panting only occurred at 35 °C. The frequency of panting was
difficult to observe because the birds often held their necks forward and down so that the
view from the small window was blocked by the bird's body. Therefore it is not possible to
look for associations between panting frequency and thermal balance. Nevertheless upon
opening the chamber at the end of the 35 °C trials, panting and an elongated snood (only
visible in uninsulated trials) were observed in all individuals. Snood elongation did not
occur at other temperatures.
Although the proportion of observations in which the subjects were standing was
not influenced by the insulation treatment, the frequency of this behavior did influence some
of the dependent thermal variables. Frequency of standing was significantly
Table 5-1. Partial results of repeated measures ANOVAs showing statistically significant
sources of variation in, and those with strong trends of influence on, the dependent
measures of thermal balance listed.
Source of Variance
Table 5-2. Mean values of dependent thermal variables in eight wild turkeys exposed to
three ambient temperatures (°C) with their bare heads and necks uinsulated or insulated.
Values presented for temperature measurements are mean changes in temperature (end of
trial value minus beginning of trial value).
Thermal Conductance (Total)
Thermal Conductance (Dry)
Core Body Temperature
Head Skin Temperature
0.7 l b
Statistically significant differences between values of uninsulated and insulated birds at the
same temperature have the following probabilities: a <0.01; b <0.05; 0.09> c >0.05.
n ■ _n
■ ■d ■■
Figure 5-1 . Oxygen consumption of eight turkeys with heads
uninsulated (empty squares) and insulated (filled squares) at three ambient
temperatures. Presented in order of increasing body mass.
■3 - iD D
■ ■ ,
n no ■
Figure 5-2. Cooling capacities of eight turkeys with heads uninsulated
(empty squares) and insulated (rilled squares) at three ambient
temperatures. Presented in order of increasing body mass.
■ ■ '
gi|E |H H H
Figure 5-3. Total thermal conductances of eight turkeys with heads
uninsulated (empty squares) and insulated (filled squares) at three ambient
temperatures. Presented in order of increasing body mass.
I I M^— ■—
» ■ -^^^^^^^^^^^^^
Figure 5-4. Change in core body temperature of eight turkeys with heads
uninsulated (empty squares) and insulated (filled squares) at three ambient
temperatures. Presented in order of increasing body mass.
correlated with higher head skin temperatures at °C (both insulation treatments combined,
r s = 0.54, p=0.05) and is associated with a tendency for lower rates of metabolic heat
production (r s = -0.50, p=0.07). At 22 and 35 °C, feather temperature inversely correlated
with standing frequency (r s = -0.66, p=0.01 and r s = -0.52, p= 0.05, respectively).
Head tucking occurred in eight of 16 trials by six different individuals at °C.
Higher frequencies of head tucking were positively, though only nearly significantly,
correlated with changes in feather and dewlap temperatures (r s = 0.49, p=0.08 and r s = 0.50,
p=0.05, respectively), and negatively correlated with changes in skin temperature (rho=
-0.58, p=0.04), changes in core body temperature (r s = -0.54, p= 0.05), metabolic heat
production (r s = -0.56, p= 0.05) and both total and dry conductance (r s = -0.56, p= 0.05 and
r s = -0.66, p= 0.02, respectively).
A dramatic cost of insulated heads and necks occurs in male wild turkeys at high
temperatures. Insulated males had higher metabolic rates and markedly increased core body
temperatures. Associated with this were increased dry and evaporative thermal
conductances over uninsulated males at the same temperature. However the much lower
cooling capacities of insulated males reveals their relative inability to dissipate metabolic
heat by evaporative heat loss. These results demonstrate that the unfeathered heads and
necks of male wild turkeys, and possibly the fleshy structures on the head, contribute to heat
dissipation at high ambient temperatures.
Contrary to expectations, under cold conditions head and neck insulation did not
significantly reduce thermal conductance or increase metabolic heat production. Under cold
conditions free living wild turkeys often contract the skin at the back of their necks,
effectively drawing the feathered skin at the base of the neck up and over much of the
usually bare areas of the back of the neck (personal observation). The captive wild turkeys
in this study exhibited similar behavior, possibly explaining the absence of a difference in
thermal conductance between uninsulated and insulated birds at °C. Because winter
starvation can be an important source of mortality for turkey populations in the northern
part of their distribution (Healy 1992b), reducing heat loss from the head may enhance
turkey survivorship. At night thermal conductance may be further decreased by tucking the
head under the wing or back feathers. In this study four of the eight uninsulated individuals
at °C were seen to have their heads tucked during at least one of the observation periods.
Three of these had lower metabolic rates than the remaining individuals, providing support
for this explanation. LaRochelle et al. (1982) found a similar effect of head tucking in black
vultures, which also have unfeathered heads. Additional studies of the effect of artificial and
behavioral head insulation on heat production and loss in wild turkeys are needed at low
Other birds have modified unfeathered areas to control heat loss. Ptarmigan
(Lagopus spp.), which live at high latitudes and altitudes where the difference between body
temperature and ambient temperature can be large (e.g., > 60 °C), often have legs and feet
that are feathered (Johnsgard 1983). Other species limit heat loss in cold conditions with
vascular modifications. Gulls (Laridae) have counter-current heat exchange mechanisms
that reduce heat loss from the feet under cold conditions (Baudinette et al. 1976). The wood
stork (Mycteria americana) and turkey vulture (Cathartes aura ) use their unfeathered legs to
dump heat at hot temperatures and are able to enhance this mechanism of heat loss by
defecating on their legs to promote evaporative heat loss (Kahl 1963; Hatch 1970). Ducks
may utilize the large surface area of their bills to dissipate heat (Hagan & Heath 1980). The
wild turkeys used in this study are the only species in which the value of unfeathered heads
and necks for heat dissipation has been demonstrated experimentally.
Previous studies of the metabolism of the wild turkey have ignored the metabolism
of wild turkeys at temperatures above 25 °C (Gray & Price 1988; Oberlag et al. 1990). The
adaptive benefit of unfeathered heads demonstrated here suggests that peak effective
temperatures during the reproductive season, especially in habitats without shade, may limit
wild turkey distribution or population density. These results are reinforced by early studies
on the temperature requirements of domestic turkeys. High ambient temperature
(approximately 30 °C) and exposure to direct sunlight may reduce male fertility by as much
as 10% in Broad-Breasted Bronze turkeys, the domestic breed most similar in appearance to
wild turkeys (Kosin & Mitchell 1955). Wilson and Woodard (1955) found that all
domestic turkeys were subject to hyperthermia at ambient temperatures above 32 °C, but
that this was particularly true of large males. In addition body temperature and water
consumption by domestic turkeys were inversely correlated with the percent shade cover
provided at ambient temperatures above 35 °C (Wilson et al. 1955; Wilson & Woodard
1955). Wild turkeys experienced heat stress at 35 °C in the lab in this study. All males
responded to hot chamber temperatures by panting, dropping their wings, and extending
their necks and snoods. One individual (no. 1) even became frantic at the very end of both
high temperature trials and was removed immediately. Behavioral changes that occur in
free-living wild males under hot conditions provide similar support.
Mature male turkeys in northern Florida seem to avoid bright sun and in the
summer are found standing in heavy shade with dewlap and neck bright red and extended
while panting heavily (personal observation). Also it seems that males are more reluctant to
flee under these conditions and can be approached more closely than when it is cooler.
Males are faced with a thermoregulatory quandry under hot, sunny conditions. Resting
quietly in the shade maintains sublethal body temperatures but does not allow feeding,
fighting for access to mates, or displaying to females. These latter activities, however, are
also functionally and adaptively necessary, but result in metabolic heat production and
exposure to solar radiation. Field studies of the behavior of wild turkeys relative to
environmental conditions, including radiative heat load and wind speed, are needed to
understand how males tradeoff thermal needs with feeding and mating success. The results
of this study suggest that the bare head and neck of male wild turkeys enables wild turkeys
to manage these conflicting goals more successfully.
The unfeathered, ornamented head of male wild turkeys has dual adaptive functions:
it indirectly increases mating success and directly increases heat dissipation at high
temperatures. The FS of wild turkeys increases mating success in two ways. Male snood
length, is strongly correlated with both female choice and male dominance. A negative
correlation between snood length and number of coccidia oocysts found in the feces
suggests that females use male snood length to detect which males are resistant to this
debilitating parasite. Females that mate with parasite-resistant males may have increased
fitness if the male's parasite resistance is heritable, because her offspring will be better able
to withstand the deleterious effects of parasitism. Longer snooded males were dominant
over shorter snooded ones and all males avoided interaction with longer snooded artificial
males. Males may avoid fighting with longer snooded males because of the risks of
combatting males that are likely to be in better overall condition.
In addition to these sexually selected functions for some aspects of the unfeathered
head of wild turkeys, FS are used in thermoregulation. Turkeys whose head and
ornamentation had been insulated as though covered with feathers were unable to dissipate
heat at high temperatures. Although the thermoregulatory function of the uninsulated head
was not experimentally attributed to any one part of the male's head ornamentation in this
study, the distention of snoods by males at high temperatures suggest that this character
may be used to dissipate heat.
The mystery of FS maintenance and evolution in wild turkeys and other galliforms
was not fully explained by the study of sexual selection and thermoregulation described
above. The head ornamentation of wild turkeys has several components, only one of which,
the snood, seemed to be under directional sexual and natural selection. In particular the
functions of the distinctive side and front caruncles remain unexplained. These characters
showed considerable variation that could be used by females and other males to discriminate
among males. Why are these characters maintained? Multiple year, mate choice studies
have shown that the characters that females seem to be assessing can vary somewhat from
year to year. It is possible that mate choice may still be an important explanation for these
characters. Additional studies are needed to address variation in character assessment over
the lifetime of both males and females in this species.
Side and frontal caruncles are highly vascularized structures that typically turn
bright red during display due to increased blood flow. My experimental treatment did not
selectively insulate these structures, but the increased surface area provided by these
characters could serve to increase the dissipation of excess body heat over that dissipated by
the plain bare skin. Thus caruncles may play a role in heat dissipation during male
courtship display. It may be possible to remove caruncles or restrict blood flow to them to
study their use in heat dissipation. Experimental studies such as this, coupled with field
observations of thermoregulatory behavior patterns can clarify the reason fleshy characters
characters other than the snood are maintained.
The highly competitive nature of highly polygynous mating systems, such as the
wild turkey's, is thought to select for increased body size in males. However large body size
poses a problem for animals in hot climates because their reduced surface area to volume
ratio makes it more difficult for these species to dissipate heat to the environment As
described in the second chapter, fleshy structures are larger and more common in larger
species. This suggests that the need for large fleshy structures in some species is caused by
increased body size driven by intense sexual selection. Species living in cooler climates and
small species in any climate should not experience the dilemma posed by hyperthermia, nor
should species that have low metabolic rates. Studies of fleshy structures relative to body
size and environmental temperatures in other avian taxa are needed to test the validity of this
Studies of sexual selection rarely consider additional or alternative hypotheses for
the function of ornamental characters. Sexual selection theory generally assumes that
greater elaboration of sexually-selected traits carries with it decreased survival. The results
presented in this dissertation suggest that sexual selection may not necessarily be at odds
with natural selection. Instead sexual and natural selection may, in some cases, work in
tandem. The costs of sexually-selected ornaments are assumed to be extreme. However
attention to non-sexual hypotheses in addition to sexually-selected ones, may reveal that the
apparent benefits of extravagant characters are multiple and far outweigh the costs. Heated
debate between theoreticians on the evolutionary stability of different models of female
choice for male characteristics, has narrowed consideration of other functions for so-called
ornaments. My work suggests that a general skepticism of the ability of sexual selection to
explain everything is warranted. "A good field worker is nobody's poodle" (Grafen 1987,
LIST OF REFERENCES
Abacus Concepts Inc. 1986. Statview 512+. Calabasas, CA,
Abacus Concepts Inc. 1989. SuperANOVA. Berkeley, CA,
Alatalo, R. V., Hoglund, J. & Lundberg, A. 1991. Lekking in the black grouse: a test of
male viability. Nature (London), 352, 155-156.
Allee, W. C, Collias, N. E. & Lutherman, C. Z. 1939. Modification of the social order in
flocks of hens by the injection of testosterone proprionate. Physiol ZooL, 12, 412-439.
Andersson, M. 1986. Evolution of condition-dependent sex ornaments and mating
preferences: sexual selection based on viability differences. Evolution, 40, 804-816.
Ashford, R. W., Palmer, T. T., Ash, J. S. & Bray, R. S. 1976. Blood parasites of Ethiopian
birds: a general survey. J. Wildl. Dis., 12, 409-426.
Atkinson, C. T. & van Riper in, C. 1991. Pathogenicity and epizootiology of avian
haematozoa: Plasmodium, Leucocytozoon and Haemoproteus. In: Bird-Parasite Interactions
(Ed. by J. Loye & M. Zuk), pp. 19-48, Oxford: Oxford University Press.
Atkinson, C. T., Forrester, D. J. & Greiner, E. C. 1988. Pathogenicity of Haemoproteus
meleagris (Haemosporina: Haemoproteidae) in experimentally infected domestic turkeys. J.
Parasitol., 74, 228-239.
Balmford, A, Thomas, A L. R. & Jones, I. L. 1993. Aerodynamics and the evolution of
long tails in birds. Nature, 361, 628-631.
Basolo, A. L. 1990. Female preference predates the evolution of the sword in swordtail
fish. Science, 250, 808-810.
Baudinette, R. V., Loveridge, J. P., Wilson, K. J., Mills, C. D. & Schmidt-Neilsen, K. 1976.
Heat loss from the feet of Herring Gulls at rest and during flight. Am. J. Physiology, 230,
Beaugrand, J., Goulet, C. & Payette, D. 1991. Outcome of dyadic conflict in male green
swordtail fish, Xiphophorus helleri : effects of body size and prior dominance. Anim.
Behav., 41, 417-424.
Bennett, A F., Huey, R. B., John-Alder, H. & Nagy, K. A 1984. The parasol tail and
thermoregulatory behavior of the Cape Ground Squirrel. Xerus inauris . Physiol. Zool., 57,
Bennett, G. F. & Herman, C. M. 1976. Blood parasites of some birds from Kenya,
Tanzania and Zaire. J. Wildl. Dis., 12, 59-65.
Bennett, G. F., Peirce, M. A. & Ashford, R. W. 1993. Avian haematozoa: mortality and
pathogenicity. J. Nat ffisL, 27, 993-1001.
Booth, D. T., Clayton, D. H. & Block, B. A. 1993. Experimental demonstration of the
energetic cost of parasitism in free-ranging hosts. Proc. Royal Soc. London. B, 253, 125-
Borgia, G. 1979. Sexual selection and the evolution of mating systems. In: Sexual
Selection and Reproductive Competition in Insects (Ed. by M.S. Blum & N. A. Blum), pp.
19-80, New York: Academic Press.
Boyce, M. S. 1990. The red queen visits Sage Grouse leks. Amer. Zool., 30, 263-270.
Bressler, G. O. & Gordeuk S. Jr 1951. Effect of cecal coccidiosis on body weight, egg
production and hatchability in chickens. Poul.ScL, 30, 509-51 1.
Brodsky, L. M. 1988. Ornament size influences mating success in male rock ptarmigan.
Anim. Behav., 36, 662-667.
Buchholz, R. 1989. Singing behavior and ornamentation in the Yellow-knobbed Curassow
(Crax daubentoni) . M.S. thesis, University of Florida, Gainesville.
Buchholz, R. 1991. Older males have bigger knobs: correlates of ornamentation in two
species of curassow. Auk, 108, 53-60.
Buchholz, R. 1992. Confusing models with tests in studies of sexual selection: reply to
Jones. Auk, 109, 199-201.
Burley, N., Krantzberg, G. & Radman, P. 1982. Influence of colour-banding on the
conspecific preferences of zebra finches. Anim. Behav., 30, 444-455.
Burtt, E. H. 1986. An analysis of physical, physiological, and optical aspects of avian
coloration with emphasis on wood-warblers. Ornithological Monographs, 38, 1-126.
Calder, W. A. & King, J. R. 1974. Thermal and caloric relations of birds. In: Avian
Biology. (Ed. by D.S. Farner & J.R. King), pp. 259-413. New York: Academic Press.
Clutton-Brock, T. & Harvey, P. 1984. Comparative studies. In: Behavioural Ecology (Ed.
by J. Krebs & N. Davies), pp. 7-29. Sunderland, MA.: Sinauer Assoc.
Collias, N. E. 1943. Statistical analysis of factors which make for success in initial
encounters between hens. Am. Nat, 77, 519-538.
Crowe, T. M. 1979. Adaptive morphological variation in helmeted guineafowl Numida
meleagris and crested guineafowl Guttera pucherani . Ibis, 121, 313-320.
Crowe, T. M. & Crowe, A A. 1979. Anatomy of the vascular system of the head and neck
of the helmeted guineafowl Numida meleagris . J. Zool. (London), 188, 221-223.
Crowe, T. M. & Withers, P. C. 1979. Brain temperature regulation in Helmeted
Guineafowl. South African Journal of Science, 75, 362-365.
Davidson, W. R. & Wentworth, E. J. 1992. Population influences: diseases and parasites.
In: The Wild Turkey: Biology and Management. (Ed. by J G. Dickson), pp. 101-143.
Hanisburg, PA: Stackpole Books.
Davison, G. W. H. 1985. Avian spurs. J. Zool. (London), 206, 353-366.
Delacour, J. & Amadon, D. 1973. Curassows and related birds. New York: American
Museum of Natural History.
Dewsbury, D. A. 1990. Fathers and sons: genetic factors and social dominance in deer
mice, Peromyscus maniculatus . Anim. Behav., 39, 284-289.
Dickson, J. A. 1992. The Wild Turkey: Biology and Management. Harrisburg, PA.:
Emlen, S. T. & Oring, L. W. 1977. Ecology, sexual selection, and the evolution of mating
systems. Science, 197, 215-223.
Endler, J. A 1983. Natural and sexual selection on colour patterns in poeciliid fishes.
Environ. Biol. Fishes, 9, 173-190.
Enquist, M. & Arak, A. 1993. Selection of exaggerated male traits by female aesthetic
senses. Nature, 361, 446-448.
Fisher, R. 1958. The Genetical Theory of Natural Selection, 2nd ed. Oxford: Clarendon
Forrester, D. J. 1991. The ecology and epizootiology of avian pox and malaria in wild
turkeys. Bull. Soc.Vector Ecol., 16, 127-148.
Forrester, D. J., Hon, L. T, Jr., L. E. W. & Austin, D. H. 1974. Blood protozoa of wild
turkeys in Florida. J. Protozoology, 21, 494-497.
Forrester, D. J., Humphrey, P. P., Telford, S. R. & Jr., L. E. W. 1980. Effects of blood-
induced infections of Plasmodium hermani on domestic and wild turkey poults. J. Wildl.
Dis., 16, 237-244.
Franklin, D. & Menkhorst, P. 1988. The bare facial skin of the Regent Honeyeater.
Australian Bird Watcher, 12, 237-238.
Gjesdal, A. 1977. External markers of social rank in willow ptarmigan. Condor, 79, 279-
Grafen, A. 1987. Measuring sexual selection: why bother? In: Sexual selection: Testing
the Alternatives (Ed. by J. Bradbury & M. Andersson), pp. 221-233. New York: J Wiley
Grafen, A. 1990a. Biological signals as handicaps. J. Theor. Biol., 144, 473-516.
Grafen, A. 1990b. Sexual selection unhandicapped by the Fisher process. J. Theor. Biol.,
Gray, B. T. & Price, H. H. 1988. Basal metabolism and energetic cost of metabolism in
wild turkeys. J. Wildl. Manag., 52, 133-137.
Greiner, E. C, Bennett, G. F., White, E. M. & Coombs, R. F. 1975. Distribution of the
avian haernatozoa of North America. Can. J. ZooL, 53, 1762-1787.
Griffen Productions 1990. America's Wild Turkey. Video. Rhinebeck, NY,
Hagan, A. A. & Heath, J. E. 1980. Regulation of heat loss in the duck by vasomotion in
the bill. J. Therm. Biol., 5,95-101.
Hamilton, W. D. & Zuk, M. 1982. Heritable true fitness and bright birds: a role for
parasites? Science, 218, 384-387.
Harvey, P. H. & Pagel, M. D. 1991. The Comparative Method in Evolutionary Biology.
Oxford: Oxford University Press.
Hatch, D. E. 1970. Energy conserving and heat dissipation mechanisms of the Turkey
Vulture. Auk, 87, 111-124.
Healy, W. M. 1992a. Behavior. In: The Wild Turkey: Biology and Management (Ed. by J
G. Dickson), pp. 46-65. Harrisburg, PA: Stackpole Books.
Healy, W. M. 1992b. Population influences: environment In: The Wild Turkey: Biology
and Management (Ed. by J G. Dickson), pp. 129-143. Harrisburg, PA: Stackpole Books.
Heisler, L., Andersson, M. B., Arnold, A. J., Boake, C. R, Borgia, G., Hausfater, G.,
Kirkpatrick, M., Lande, R., Smith, J. M., O'Donald, P., Thornhill, A. R & Wessing, F. J.
1987. The evolution of mating preferences and sexually selected traits. In: Sexual
Selection: Testing the Alternatives (Ed. by J. Bradbury & M. Andersson), pp. 97-1 18. NY:
J Wiley & Sons.
Henneman, W. I. 1988. Energetics and spread-winged behavior in Anhingas and Double-
Crested Comorants: risks of generalization. Amer. Zool., 28, 845-851.
Hill, G. E. 1994. Geographic variation in male ornamentation and female mate preference
in the house finch: a comparative test of models of sexual selection. Behav. Ecol., 5, 64-73.
Hill, R., Beaver, D. & Veghte, J. 1980. Body surface temperatures and thermoregulation in
Black-capped Chickadees (Parus atricapillus) . Physiol. ZooL, 53, 305-321.
Hillgarth, N. 1990. Parasites and female choice in the Ring-necked Pheasant. Amer. Zool.,
Holder, K. & Montgomerie, R. 1993. Context and consequences of comb displays by
male rock ptarmigan. Anim. Behav., 45, 457-470.
Hollett, K., Thomas, V. & MacDonald, S. 1984. Structural and functional aspects of
supraorbital combs of grouse. International Grouse Symposium, York, 3, 193-21 1.
Hon, L. T., Forrester, D. J. & Williams, L. E. Jr 1975. Helminths of wild turkeys in
Florida. Proc. Helminthol. Soc.Wash., 42, 119-127.
Hon, L. T., Forrester, D. J. & Williams, L. E. Jr 1978. Helminth acquisition by wild
turkeys (Meleagris gallopavo osceola) in Florida. Proc. Helminthol. Soc.Wash., 45, 211-
Hudson, P. J. & Dobson, A. P. 1991. The direct and indirect effects of the caecal
nematode Trichostrongylus tenuis on red grouse. In: Bird-Parasite Interactions (Ed. by J.
Loye & M. Zuk), pp. 49-68. New York: Oxford University Press.
Hurst, G. A. 1992. Foods and Feeding. In:The Wild Turkey: Biology and Management.
(Ed. by J.G. Dickson), pp. 66-83. Harrisburg, PA: Stackpole Books.
Johnsgard, P. 1983. The Grouse of the World. Lincoln: University of Nebraska Press.
Johnsgard, P. 1986. The Pheasants of the World. Oxford: Oxford University Press.
Johnsgard, P. 1988. The Quails, Partridges and Francolins of the World. Oxford: Oxford
Johnson, K., Dalton, R. & Burley, N. 1993. Preferences of female American goldfinches
(Carduelis tristis') for natural and artificial male traits. Behav. Ecol., 4, 138-143.
Johnson, K. & Mazluff, J. 1990. Some problems and approaches in avian mate choice.
Auk, 107, 296-304.
Johnson, L. W. & Edgar, S. A. 1982. Responses to prolonged selection for resistance and
susceptibility to acute cecal coccidiosis in the Auburn Strain Single Comb White Leghorn.
Poult. Sci., 61, 2344-2355.
Kahl, M. P. Jr 1963. Thermoregulation in the Wood Stork with special reference to the
role of the legs. Physiol. Zool., 36, 141-151.
Kelly, G. 1975. Indexes for aging eastern wild turkeys. In: Proceedings of the Third
National Wild Turkey Symposium (Ed. by L.K. Halls), pp. 205-209. Edgefield, SC:
National Wild Turkey Federation.
Kirkpatrick, M. 1987. The evolutionary forces acting on female mating preferences in
polygynous animals. In: Sexual Selection: Testing the Alternatives (J. Bradbury & M.
Andersson), pp. 67-82. Chichester: John Wiley & Sons.
Kodric-Brown, A. 1989. Dietary carotenoids and male mating success: an environmental
component to female choice. Behav. Ecol. Sociobiol., 25, 393-401.
Kodric-Brown, A. 1990. Mechanisms of sexual selection: insights from fishes. Ann. Zool.
Fennici, 27, 87-100.
Kodric-Brown, A. 1993. Female choice of multiple male criteria in guppies: interacting
effects of dominance, coloration and courtship. Behav. Ecol. Sociobiol., 32, 415-420.
Kodric-Brown, A & Brown, J. 1984. Truth in advertising: the kinds of traits favored by
sexual selection. Am. Nat, 124, 309-323.
Korpimaki, E., Hakkarainen, H. & Bennett, G. F. 1993. Blood parasites and reproductive
success of Tengmalm's owls: detrimental effects on females but not males? FuncL Ecol., 7,
Kosin, I. L. & Mitchell, M. S. 1955. Ambient temperature as a factor in turkey
reproduction. Poul. Sci., 34, 499-505.
Kruijt, J. P. 1964. Ontogeny of social behavior in Burmese red jungle fowl (Gallus gallus
spadiceus) . Behav. SuppL, 12, 1-201.
LaRochelle, J., Delson, J. & Schmidt-Nielsen, K. 1982. Temperature regulation in the
Black Vulture. Can. J. Zool., 60, 491-494.
Ligon, J., Thornhill, R., Zuk, M. & Johnson, K. 1990. Male-male competition,
ornamentation, and the role of testosterone in sexual selection in red jungle fowl. Anim.
Behav., 40, 367-373.
Lisano, M. E. & Kennamer, J. E. 1977. Seasonal variations in the plasma testosterone level
in male eastern wild turkeys. J. Wildl. Manag., 41, 184-188.
Lucas, A. & Stettenheim, P. 1972. Avian Anatomy: Integument Washington, DC: USDA
Manning, J. 1985. Choosy females and correlates of male age. J. Theor. BioL, 116, 349-
Manning, J. 1989. Age advertisement and the evolution of the peacock's train. J. Evol.
Biol., 2, 299-313.
Martin, P. & Bateson, P. 1986. Measuring Behaviour. Cambridge: Cambridge University
McClure, H. E., Poonswad, P., Greiner, E. C. & Laird, M. 1978. Haematozoa in the Birds
of Eastern and Southern Asia. St. John's, Newfoundland: International Reference Centre for
McKinney, F., Sorenson, L. G. & Hart, M. 1990. Multiple functions of courtship displays
in dabbling ducks (Anatini). Auk, 188-191.
McNab, B. K. 1988. Fn^rgy conservation in a tree-kangaroo (Dendrolagus matschiei) and
the red panda (Ailurus fulgens). Physiol. Zool., 61, 280-292.
Miller, J. E. & Leopold, B. D. 1992. Population influences: predators. In: The Wild
Turkey: Biology and Management (Ed. by J. G. Dickson), pp. 1 19-128. Harrisburg, PA:
M0ller, A. P. 1987. Social control of deception among status signalling house sparrows
Passer domesticus . Behav. Ecol. Sociobiol., 20, 307-311.
Morony, J. J. Jr., Bock, W.J. & Farrand, J. Jr. 1975. Reference list of the birds of the
world. DepL of Ornithology, American Museum of Natural History, New York.
Moss, R., Kolb, H., Marquiss, M., Watson, A., Treca, B., Watt, D. & Glennie, W. 1979.
Aggressiveness and dominance in captive red grouse. Aggressive Behavior, 5, 59-84.
Oberlag, D. F., Pekins, P. J. & Mautz, W. W. 1990. Influence of seasonal temperatures on
wild turkey metabolism. J. Wildl. Manag., 54, 663-667.
Ogbuokiri, U. D. E. & Edgar, S. A. 1986. Effect of infections with Eimeria spp. on sperm
production of broiler strain sires. Poult. Sci., 65, 184-185.
Oosthuizen, J. H. & Markus, M. B. 1967. The Haematozoa of South African birds. I:
Blood and other parasites of two species of game birds. Ibis, 109, 115-117.
Oosthuizen, J. H. & Markus, M. B. 1969. The haematozoa of South African birds. V:
Report on blood smears collected by the September 1967 Percy FitzPatrick Institute
expedition to the Chobe. Ostrich, 40, 21-22.
Pagel, M. & Harvey, P. 1988. Recent developments in the analysis of comparative data. Q.
Rev. Biol., 63, 413-440.
Parker, G. A. 1974. Assessment strategy and the evolution of fighting behaviour. J. Theor.
Payne, R. B. 1984. Sexual selection, lek and arena behavior, and sexual size dimorphism in
birds. Ornithol. Monog., 33, 1-52.
Peirce, M. A. 198 1 . Distribution and host-parasite checklist of the hematozoa of birds in
Western Europe. J. Nat. Hist, 15, 419-458.
Pelham, P. H. & Dickson, J. G. 1992. Physical Characteristics. In:The Wild Turkey:
Biology and Management (Ed. by J.G. Dickson), pp. 32-45. Harrisburg, PA: Stackpole
Powell, J.A. 1965. The Florida Wild Turkey. Tech. Bull. No. 8, Florida Game and Fresh
Water Fish Commission, Tallahassee, FL.
Prestwood, A. K., Kellogg, F. E. & Doster, G. L. 1973. Parasitism and disease among
southeastern wild turkeys. In: Wild Turkey Management: Current Problems and Programs
(Ed. by G.C. Sanderson & H.C. Schultz), pp. 159-167. Columbia, MO: University of
Randi, E., Fusco, G., Lorenzini, R. & Crowe, T. 1991. Phylogenetic relationships and rates
of allozyme evolution within the Phasianidae. Biochem. Syst Ecol. 19, 213-221.
Read, A. 1987. Comparative evidence supports the Hamilton and Zuk hypothesis on
parasites and sexual selection. Nature, 328, 68-70.
Reichholf, J. 1988. Macht der Sperbergeier Gyps rue ppellii ein "Drohgesicht". Verhalten
Ornithologische Geselschaft Bayern, 24, 751-755.
Richner, H. 1989. Phenotypic correlates of dominance in carrion crows and their effects
on access to food. Anim. Behav., 38, 606-612.
Rohwer, S. 1975. The social significance of avian winter plumage variability. Evolution, 29,
Rohwer, S. 1977. Status signalling in Harris 1 sparrows: some experiments in deception.
Behaviour, 61, 107-129.
Ruff, M. 1988. Effect of coccidiosis on reproductive maturation of male Japanese Quail.
Avian Dis., 32, 41-45.
Ruff, M. D., Fagan, J. M. & Dick, J. W. 1984. Pathogenicity of coccidia in Japanese quail
(Coturnix coturnix japonica) . Poul. ScL 63, 55-60.
Ruff, M. D., Reid, W. M. & Johnson, J. K. 1974. Lowered blood carotenoid levels in
chickens infected with coccidia. Poul. Sci., 53, 1801-1809.
Ruff, M. D., Schorr, L., Davidson, W. R. & Nettles, V. F. 1988. Prevalence and identity of
coccidia in pen-raised wild turkeys. J. Wildl. Manag., 24, 71 1-714.
Ruff, M. D. & Wilkins, G. C. 1987. Pathogenicity of Fjmeria lettyje. (Ruff 1985) in the
northern bobwhite (Colinus virginianus . L.). J. Wildl. Dis., 23, 121-126.
SAS Institute Inc. 1989. JMP User's Guide Version 2. Cary, NC,
Sasseville, V. G, Miller, B. & Nielsen, S. W. 1988. A pathologic study of wild turkeys in
Connecticut. Cornell Veterinarian, 78, 353-364.
Siegel, S. 1956. Nonparametric Statistics for the Behavioral Sciences. New York:
Sokal, R. R. & Rohlf, F. J. 1981. Biometry. New York: W H Freeman and Co.
Spurrier, M. F., Boyce, M. & Manly, B. F. 1991. Effects of parasites on mate choice by
captive sage grouse (Centrocercus urophasianus) . In: Ecology, Behaviour and Evolution of
Bird-Parasite Interactions (Ed. by J. Loye & M. Zuk), pp. 389-398. Oxford: Oxford
Steffen, D. E., Couvillion, C. E. & Hurst, G. A 1990. Age determination of eastern wild
turkey gobblers. Wild. Soc. Bull., 18, 119-124.
Stokkan, K. 1979. Testosterone and daylength-dependent development of comb size and
breeding plumage and comb growth in male willow ptarmigan (Lagopus lagopus lagopus ).
Auk, 96, 106-115.
Sullivan, M. S. 1994. Mate choice as an information gathering process under time
constraint: implications for behaviour and signal design. Anim. Behav., 47, 141-151.
Thiessen, D. D. & Kittrell, E. M. W. 1979. Mechanical features of ultrasound emission in
the mongolian gerbil. Amer. Zool., 19, 509-512.
von Schantz, T., Goransson, G., Andersson, G., Froberg, I., Grahn, M., Helgee, A. &
Wittzell, H. 1989. Female choice selects for a viability-based male trait in pheasants.
Nature, 337, 166-169.
Watts, C. R. & Stokes, A W. 1971. The social order of turkeys. Scientific American, 224,
Welty,J.C. 1975. The Life of Birds. Philadelphia, PA: W.B. Saunders Co.
White, E., Greiner, E. C, Bennett, G. F. & Herman, C. F. 1978. Distribution of the
hematozoa of Neotropical birds. Rev. Biol. Trop., 26, 43-102.
Whittow, G. 1986. Regulation of body temperature. In: Avian Physiology (Ed. by P.D
Sturkie), pp. 221-252. New York: Springer- Verlag.
Williams, L. & Austin, D. 1988. Studies of the Wild Turkey in Florida. Gainesville:
University Presses of Florida.
Wilson, W. O., Edwards, W. H., Plaister, T. L., Hillerman, J. & Woodard, A. 1955. The
shade requirement of growing turkeys. Poul. Sci., 34, 505-508.
Wilson, W. O. & Woodard, A. 1955. Some factors affecting body temperatures of
turkeys. Poul. Sci., 34, 369-37 1 .
Wingfield, J. C, Ball, G. F., Dufty, A M. Jr, Hegner, R. E. & Ramenofsky, M. 1987.
Testosterone and aggression in birds. Am. Sci., 75, 602-608
Wink, M. & Bennett, G. F. 1976. Blood parasites of some birds from Ghana. J. Wildl.
Dis., 12, 587-590.
Withers, P. & Crowe, T. 1980. Brain temperature fluctuations in helmeted guineafowl
under semi-natural conditions. Condor, 82, 99-100.
Zahavi, A. 1975. Mate selection- a selection for a handicap. J. Theor. Biol., 53, 205-214.
Zuk, M., Johnson, K., Thornhill, R. & Ligon, J. 1990a. Mechanisms of female choice in
red jungle fowl. Evolution, 44, 477-485.
Zuk, M., Ligon, J. D. & Thornhill, R. 1992. Effects of experimental manipulation of male
secondary sex characters on female mate preference in red jungle fowl. Anim. Behav., 44,
Zuk, M., Thornhill, R., Ligon, J. & Johnson, K. 1990b. Parasites and mate choice in red
jungle fowl. Amer. Zool., 30, 235-244.
Richard Buchholz was born in Manhasset, New York, on 5 November 1964 and
grew up in nearby Little Neck. His early interest in biology was satisfied by wandering the
salt marshes and seashore along Little Neck Bay on Long Island Sound and exploring the
forest during summer trips to his uncle's home in Mountaintop, Pennsylvania.
He majored in biology at the State University of New York at Binghamton and
graduated with a Bachelor of Science degree, with outstanding overall performance and
distinguished independent study in biological sciences, in June 1986. Subsequently he
studied the reproductive behavior of the yellow-knobbed curassow in the wild in Venezuela
and in captivity in Mexico for his Master of Science degree, which was granted by the
University of Florida, Department of Zoology, in August 1989. This work led to his
interest in the costs and benefits of wattles, ceres and similar structures in birds. He
received his Doctor of Philosophy degree from the Department of Zoology, University of
Florida in August 1994.
I certify that I have read this study and that in my opinion it conforms to acceptable
l„Jl%i^%**** anally adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
H: Jane Brockmanrv Chairman
Professor of Zoology
I certify that I have read this study and that in my opinion it conforms to acceptable
stand JSoSly P^entation and i fully adequate, in scope and quahty, as a
.• ._*:„„ e^iu* Aoarpt* of Doctor of Philosophy.
stanaaras 01 sunui<ury V i^o^^~~ --., ' . /
dissertation for the degree of Doctor of Philosophy.
Richard A. Kiltie
Associate Professor of Zoology
I certify that I have read this study and that in my opinion it conforms to acceptable
sm J^loXv^^iion and is fully adequate, in scope and quahty, as a
dissertation for the degree of Doctor of Philosophy.
uougiaj^a. j-^vt-jr /
Associate Professor of Zoology
I certify that I have read this study and that in my opinion it conforms to acceptable
stand Js ^scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
Professor of Zoology
I certify that I have read this study and that in my opinion it ™^ to J?fP table
stand Js of sc'holarly presentation and Lis fully adequate, In scope an^fty, as a
dissertation for the degree of Doctor of Philosophy.
Jllis C. Greiner
Professor of Veterinary Medicine
7 S&aJI<<iJSm - j&dri
Dean, Graduate School